Clinical Oncology (2007) 19: 397e417 doi:10.1016/j.clon.2007.03.010

Overview

Chemical Radiosensitizers for Use in Radiotherapy*

P. Wardman

University of Oxford, Gray Cancer Institute, PO Box 100, Mount Vernon Hospital, Northwood HA6 2JR, UK

ABSTRACT: Radiosensitizers are intended to enhance tumour cell killing while having much less effect on normal tissues. Some drugs target different physiological characteristics of the tumour, particularly hypoxia associated with radioresistance. Oxygen is the definitive hypoxic cell radiosensitizer, the large differential radiosensitivity of oxic vs hypoxic cells being an attractive factor. The combination of nicotinamide to reduce acute hypoxia with normobaric carbogen breathing is showing clinical promise. ‘Electron-affinic’ chemicals that react with DNA free radicals have the potential for universal activity to combat hypoxia-associated radioresistance; a nitroimidazole, nimorazole, is clinically effective at tolerable doses. Hypoxia-specific cytotoxins, such as tirapazamine, are valuable adjuncts to radiotherapy. Nitric oxide is a potent hypoxic cell radiosensitizer; variations in endogenous levels might have prognostic significance, and routes to deliver nitric oxide specifically to tumours are being developed. In principle, many drugs can be delivered selectively to hypoxic tumours using either reductase enzymes or radiation-produced free radicals to activate drug release from electron-affinic prodrugs. A redox-active agent based on a gadolinium chelate is being evaluated clinically. Pyrimidines substituted with bromine or iodine are incorporated into DNA and enhance free radical damage; fluoropyrimidines act by different mechanisms. A wide variety of drugs that influence the nature or repair of DNA damage are being evaluated in conjunction with radiation; it is often difficult to define the mechanisms underlying chemoradiation regimens. Drugs being evaluated include topoisomerase inhibitors (e.g. camptothecin, topotecan), and the hypoxia-activated anthraquinone AQ4N; alkylating agents include temozolomide. Drugs involved in DNA repair pathways being investigated include the potent poly(ADP ribose)polymerase inhibitor, AG14361. Proteins involved in cell signalling, such as the Ras family, are attractive targets linked to radioresistance, as are epidermal growth factor receptors and linked kinases (drugs including vandetanib [ZD6474], cetuximab and gefitinib), and cyclo- oxygenase-2 (celecoxib). The suppression of radioprotective thiols seems to offer more potential with alkylating agents than with radiotherapy, although it remains a strategy worthy of exploration. Wardman, P. (2007). Clinical Oncology 19, 397e417 ª 2007 The Royal College of Radiologists. Published by Elsevier Ltd. All rights reserved.

Key words: Chemoradiation, DNA, hypoxia, nitroimidazoles, oxygen, radiosensitizers

Introduction an example d can react with these free radicals and modify response. Differences in radiosensitivity might In the present context, the measures of radiosensitivity of reflect variations in the levels or activity of proteins most interest are the clonogenic survival of tumour cells, involved in the repair of damage to DNA, linked in turn to and the survival of cells in, or functionality of, normal gene expression: chemicals that inactivate such proteins tissues, after doses of radiation delivered with therapeutic might be radiosensitizers. As cells progress (or not) through intent. Variations in these measures of radiosensitivity the cell cycle, checkpoints and signalling events may vary reflect many factors. Differences in response with radiation in their efficiencies, and can be modified by drugs. quality might arise from different distributions of the initial We consider here only the modulation of radiosensitivity ionization events, leading to differences in the nature, by low molecular weight chemicals. These can be both yields and/or spatial distribution (especially clustering) of endogenous substances, such as oxygen, nitric oxide, thiols damage from the free radicals that are the ultimate cause and ascorbate (the levels of all of which can both vary and of cell death or pathological change. Chemicals d oxygen is be modulated), and xenobiotic chemicals, which interact with radiation damage in some way. They can be further separated into substances that react with short-lived free * The chemical structures of many of the drugs referred to in radicals and need to be present at the instant of irradiation this overview are shown in Fig. 1; these are asterisked on first (e.g. oxygen), and those that target radiation effects more mention. indirectly, such as by binding to DNA repair enzymes or cell

0936-6555/07/190397þ21 $35.00/0 ª 2007 The Royal College of Radiologists. Published by Elsevier Ltd. All rights reserved. 398 CLINICAL ONCOLOGY signalling proteins to render them ineffective. Radiation Types of Chemical Radiosensitizer therapy is often given in conjunction with a course of chemotherapy; in some instances this includes regimens in An early pioneer in this field, G. E. Adams, divided which therapeutic gain is sought by exploiting synergy radiosensitizers into five categories [15,16]: between radiation and drug effects. An example would be the combination of drugs that kill radioresistant hypoxic ‘Suppression of intracellular-SH [thiols] or other endo- cells with a radiotherapy course. Although the planning of genous radioprotective substances. such a regimen would consider the two treatments in Radiation-induced formation of cytotoxic substances concert because the target cell population varies during the from the radiolysis of the sensitizer. radiotherapy course because of differential radiosensitivity Inhibitors of post-irradiation cellular repair processes. of hypoxic vs oxic cells, the effects are in principle Sensitization by structural incorporation of thymine independent, and this topic is not discussed here. However, analogues into intracellular DNA. some drugs may both kill hypoxic cells and react with short- Oxygen-mimetic sensitizers, for example the electron- lived, radiation-produced free radicals; these are discussed affinic drugs .’. below. Furthermore, the independence of action is often a grey area: if one defines radiation effects to include All these types of radiosensitizer are discussed below, a long post-irradiation period, it may be unclear whether although the order and emphasis is changed, and there is any effects of chemotherapy given any time after irradia- new interest in cell signalling processes and growth factors tion are truly independent of radiobiological effects. The so that post-irradiation pathways of interest extend beyond terminology in discussing the interaction of cytotoxic DNA repair. chemotherapy with radiation has long been a problem Another leader in this area, E. J. Hall, in discussing [1,2], yet ‘our understanding of the specific mechanisms of radiosensitizers, stressed the importance of a differential interaction between radiation and chemotherapy is still effect between tumours and normal tissue, and with this evolving’ [3]. The present overview cannot hope to ‘all important criterion’ suggested in the fifth edition of his encompass all aspects of the interaction of radiation with standard text [17] ‘only two types of sensitizers have found drugs; a recent review [4] set out the general principles of practical use in clinical radiotherapy: the ‘concurrent chemoradiation paradigm’, and previous papers have discussed the biological basis for combining The halogenated pyrimidines . based on the premise drugs with radiation [3] and reviewed many trials of that tumor cells cycle faster and, therefore, incorpo- combined radiation/drug treatment [5]. A comprehensive rate more drug than the surrounding normal tissues. overview of both radiation sensitizers and protectors Hypoxic cell sensitizers increase the sensitivity of cells showed the breadth of the topic [6]; new and emerging deficient in molecular oxygen . based on the premise radiosensitizers and radioprotectors have been reviewed that hypoxic cells occur only in tumors and not in recently [7], focusing on the newer chemoradiation normal tissues.’ modalities. A report of a meeting to advise the International Atomic Energy Agency on radiosensitizers meriting further This focus now seems too narrow to the present author, or development also described the newer approaches [8]. at least ‘hypoxic cell sensitizers’ can now be broadened Chemical radioprotectors are, of course, the reverse of as a term far beyond the original concept. Taking up the radiosensitizers: the aim is to decrease radiosensitivity, latter premise presented by Hall (recognising that the especially of normal tissues. Clinical gain can be either by issue is not clear-cut, with a spectrum of oxygen tensions a reduction in morbidity if the effects are confined to normal across both tumours and normal tissues), it is pertinent tissues, or by exploiting the hoped-for reduced radiosensi- to note recent progress in oxygen-sensitive drug delivery. tivity of normal tissues to deliver higher radiation doses and, In principle, many drugs can be specifically released only thus, enhanced tumour cell kill, the latter strategy obviously in cells of low, defined oxygen tension, exploiting the not without risk. The best-known radioprotector is the thiol ‘trigger-effector’ concept, developed especially by the prodrug, amifostine* (WR-2721). Activity in this field has group of W. A. Denny and W. R. Wilson [18], and now been included in other reviews [6,7,9e12] and is not attracting wider attention [19,20]. This approach, illus- discussed in detail here. The importance of chemical radio- trated in Fig. 2, involves constructing prodrugs comprising protectors is that their existence illustrates the competition a bioreducible ‘trigger’ (often nitroaromatic moieties between the enhancement of damage (e.g. by oxygen or drugs) based on experience with ‘electron-affinic’ radiosensi- and ‘repair’ in the specific example involving the reaction tizers), which when reduced by cellular enzymes (donating of short-lived free radicals with thiols, or thiol drugs [6,9]. an electron to form a radical anion), fragments to release As key discoveries in the 1970s relevant to this area are active drug. This release can be made selective to hypoxia becoming less well known with time (an example is the because the intermediate prodrug radical is oxygen millisecond timescale of the ‘oxygen effect’ [13,14]), some reactive; oxygen inhibits drug release via a fast, free early landmark advances are noted, along with a brief radical (electron transfer) reaction. Profiling drug release overview of the current status. The field is too large for to oxygen tensions involves matching the rates (chemical a comprehensive survey in this overview, and only kinetics) of the reactions involved [21]. A recent illustra- illustrative references are given. tion from the author’s institute shows significant CHEMICAL RADIOSENSITIZERS IN RADIOTHERAPY 399

O H O O O N CH3 CH3 H O CO2H P N H2N N OCH3 H C N S OH H C O 3 OH CH 3 H3C 3 H C Cl 3 O N N amifostine OH clofibrate efaproxiral CH3 O pentoxifylline O OH CH OCH 3 3 N OCH3 OH N OH N N H NO2 HN N(CH3)2 N N H N O N O N CH NO O2N 2 3 2 N N N NO2 NO2 N N N N N N metronidazole misonidazole nimorazole sanazole NO2 N etanidazole pimonidazole O nitracrine O – O– OH OH CH H 3 N HN O N+ N N + N Br OH O HN CH3 F CH 3 NH + N N N 2 N NH2 NO2 NO2 O – O N N O N N NH CH3 N 2 O N tirapazamine CH3 RSU1069 RB6145 OH O HN NN CH3 N CH3 O N + AQ4N O OH OH O– NH HO O N O temozolomide OH O CH O capecitabine N 3 O F N O N N S H3C N N O H N OH F camptothecin 2 N O N FBr N gemcitabine N H C HO O CF lisofylline CH3 3 3 H CO N Cl HN 3 AG14361 N(CH3)2 H CO 3 N H3C celecoxib NN N N O ONH O Cl NH N O N H CSCO H O 3 2 N NC N L778,123 H3C vandetanib gefitinib F buthionine sulfoximine NH2 Fig. 1 e Some of the drugs discussed in the text (those asterisked at first mention).

progress in designing prodrugs having the desired charac- Radiotherapy is Free Radical Therapy: the teristics [22]. Basis for Oxygen and ‘Oxygen-mimetic’ Hence, it is, in principle, possible to deliver any drug that enhances cellular radiosensitivity to tissues having low Hypoxic Cell Radiosensitizers oxygen tensions. The inclusion by Adams in his 1973 With developments in cell biology, effort in radiobiology classification of ‘Inhibitors of post-irradiation cellular repair research has shifted down the radiation effects’ timeline: processes’ is especially relevant today because of the one review a few years ago entitled ‘How does radiation kill vastly increased understanding of DNA repair and cell cells?’ does not refer to free radicals at all [23], reflecting signalling, and the development of potent inhibitors of interest in targeting cell signalling pathways as an emerging these pathways. Such inhibitors need not focus on a funda- and attractive therapeutic strategy. However, most insight mental differential between pathways in tumours vs normal into the use of arguably the most important class of chemical tissues if drug delivery can be made tumour specific, e.g. by radiosensitizers can be gained by recognising that short-lived hypoxia-controlled targeting via oxygen-sensitive prodrug free radicals are obligate intermediates in the complex fragmentation. Of course, the pharmacological and phar- pathways leading ultimately to cell kill after radiation: macodynamic characteristics of prodrug and drug need to be ‘ionizing radiation’ implies free radical production d ionisa- such that desirable ‘bystander’ effects in oxygenated tion is loss of an electron and free radicals are species with tumour cells are obtained without systemic redistribution unpaired electrons. Oxygen, the prototypical radiosensitizer of drug thwarting initially specific delivery. There is in many respects, is itself a free radical, but unusual in obvious scope to apply this approach to drugs that have having two unpaired electrons and rapidly adding to many been used in conjunction with radiotherapy and that other free radicals, producing new, reactive radicals. modulate DNA repair or cell signalling pathways. 400 CLINICAL ONCOLOGY

inactive prodrug amount effector of active linker trigger drug released modifier 00.51 % oxygen reductase enzymes or radicals from radiation

radical fragments radical reacts if O2 is low (hypoxic with O2 in tumour cells) normal tissues prodrug free radical

oxygen radicals + + antioxidant enzymes ‘trigger’ active drug prodrug residue released recycled detoxified Fig. 2 e Illustration of how drugs can be delivered to hypoxic tissues using the ‘triggereeffector’ concept. Inactive prodrugs can be reduced either by cellular enzymes or by radiation-produced radicals, the resulting free radical then fragmenting to release active drug or reacting with oxygen to recycle the prodrug. The competition between the two pathways is controlled by the kinetics of the reactions and oxygen tension. Chemical groups (‘modifiers’) on the linking moiety can change the profile of active drug release to suit the desired oxygen tension targeted.

Oxygen: the Definitive Hypoxic Cell Although there has been much work in the last decade in Radiosensitizer characterising the form of radiation survival curves at clinically relevant (!5 Gy) doses using in vitro models [30], The links between tumour hypoxia and prognosis after not always translating to corresponding effects in vivo [31], radiotherapy, as well as with tumour proliferation, sensi- almost all of the data refer to air-equilibrated cells (which tivity to some chemotherapy regimens, and the malignant incidentally all lack ascorbate, a key ‘radical sink’). There phenotype, are now well established [24]. Two recent is evidence that oxygen is less efficient in radiosensitizing reviews discussing hypoxia in head and neck cancer [25,26], cells at therapeutic compared with higher radiation and a commentary providing a brief overview of the current doses [32e34], but again studies generally compare air- understanding of tumour hypoxia [27] illustrate both the equilibrated cells with anoxia and not the response at diversity of hypoxia-associated phenomena and progress physiologically relevant (much lower) oxygen tensions. An towards patient profiling using diagnostic probes. The illustration of the response measured at clinically relevant simplest representation of the role of oxygen in influencing doses and oxygen tensions in one in vitro model in a recent radiosensitivity, and thiols in ‘repairing’ radiation damage, study in the author’s institute [35] is shown in Fig. 3.In / is often seen in equations such as: ‘R þ O2 damage spite of the historical data being based on high radiation fixation’, in competition with: ‘R þ thiol / damage repair’, doses, it is clear that many common tumours include where R represents an unspecified free radical. These a significant fraction of cells with intracellular oxygen correctly suggest a competing scenario between oxygen and concentrations in the steeply rising radiosensitivity re- thiols, but tell us nothing about both the mechanisms sponse/concentration region, such that any method to involved (‘fixing’ damage is particularly uninformative and improve tumour oxygenation should result in an increase in misleading) and the efficacy of oxygen, in concentration radiosensitivity of hypoxic subpopulations. The potential terms, as a chemical radiosensitizer. The latter is not easy gain of radiosensitizing hypoxic cells is large: severely to measure, even in vitro, because of cellular respiration hypoxic cells typically require two to three times and inefficient gas/liquid equilibration [28], although higher radiation doses to kill them compared with well- appropriate techniques have been developed [29]. oxygenated cells [17], a factor independent of absolute The oxygen concentration giving a response midway radiosensitivity across organisms differing in radiosensitiv- between well-oxygenated and anoxic cells (with a typical ity by orders of magnitude. maximum differential radiosensitivity of about a factor The most direct, and perhaps the earliest, treatment for e 2.5 3) is equivalent to around 0.4% v/v oxygen (gas phase) raising tumour oxygen levels is hyperbaric oxygen: clinical or 0.4 kPa (3 mm Hg) partial pressure (pO2) [17], but most trials have been well documented, most benefit being seen early measurements involved much higher radiation doses with few/large fraction radiation regimens [36,37]. The than are commonly given in fractionated radiotherapy. method is cumbersome and with some increase in morbidity CHEMICAL RADIOSENSITIZERS IN RADIOTHERAPY 401

1 mice [46], but clinical studies are awaited. Recombinant human erythropoietin has been evaluated clinically in some 0.9 N2 N2 detail as a means to correct haemoglobin levels [43], but its 0.8 50 ppm many functions have been noted, including its angiogenic 0.7 and mitogenic potential [42]. Reviewing published studies, 40 ppm Kaanders et al. [47] concluded that ‘The potential of 0.6 erythropoietin to promote tumor growth must be further 100 investigated, but it is too early to withdraw erythropoietin 0.5 1000 ppm from the clinic for this reason’. Indeed, a recent evaluation

Surviving fraction 80 ppm [48] suggested: ‘Most studies suggest that erythropoietic therapy either improves survival or has no negative effect 0.4 Oxygen Nitric oxide on survival when used to treat anaemia in patients with cancer . When used according to its licensed indications, 012345 012345 it is likely that erythropoietic therapy has no negative Dose / Gy effect on survival.’ Fig. 3 e Radiosensitization of hypoxic Chinese hamster V79-379A Clofibrate* is an anti-lipidaemic drug that reduces the fibroblast-like cells in vitro by either oxygen or nitric oxide at very affinity of haemoglobin for oxygen and thus acts as low concentrations (gas phase ppm v/v) and low radiation doses; a radiosensitizer [49], but clinical studies in this context clonogenic survival data from [35]. Anoxic (N ) (open circle); 2 do not appear to have been carried out. Efaproxiral* 50 ppm O2 (solid circle); 100 ppm O2 (solid square); 1000 ppm O2 (solid up triangle); 40 ppm nitric oxide (solid down triangle); (RSR13) also interacts with haemoglobin in a non-covalent 80 ppm nitric oxide (solid diamond). Lines are drawn to guide the and allosteric manner to lower the oxygen-binding affinity, eye; the dashed line is the response in anoxia (data points are and increase pO2 [50e52]. A number of clinical trials are in omitted from the left-hand panel for clarity). Although cellular progress or have been completed [46,53e55]. A phase III consumption might cause the efficacy of oxygen to be under- trial of efaproxiral, involving 515 patients treated with estimated, nitric oxide seems to be more potent than oxygen (NB whole-brain radiotherapy for brain metastases, showed nitric oxide is w50% more soluble in media than oxygen for the a significant increase in response rate for the efaproxiral same partial pressure). arm at 3 and 6 months, with evidence that patients with breast primary tumours responded better [55]. Pentoxifylline* has been reported to improve tumour in some sites [36]. Although hyperbaric oxygen is not in use oxygenation and improve radiation response if given before today with radiation, its use to treat late radiation damage irradiation in several animal studies (e.g. [56]); its modes of is attracting attention [38]. Oxygen carriers, mainly based action include increased red cell deformability, reduced on enhanced solubility parameters of perfluorocarbons blood viscosity as a result of decreasing platelet aggrega- compared with blood plasma, have been evaluated, with tion, and vasodilation. Reviewing the evidence, however, mixed results; thus one recent animal study [39] found Nieder et al. [57] considered ‘. these findings have not benefit only when the perfluorocarbon was combined with translated into positive clinical studies to date. None of carbogen breathing (carbogen is usually 95e98% O2 þ 5e2% three published clinical trials attempting to enhance the CO2 v/v). However, newer approaches are being developed, effectiveness of radiotherapy with Ptx [pentoxifylline] had exploiting advances in nanotechnology [40], and there are a satisfactory design’. reasonable prospects for the development of new oxygen Nicotinamide was originally evaluated in radiobiology as carriers that might have value in radiotherapy. an inhibitor of DNA repair via its interaction with poly(ADP Other approaches to increase the oxygen supply to the ribose) polymerase (PARP), as an analogue of known PARP tumour have been attempted; strategies to modify oxygen inhibitors, but has found more value as a vasoactive agent. transport via haemoglobin binding look particularly in- It showed good activity in combination with carbogen or teresting. The effect of anaemia on radiation therapy has normobaric oxygen with fractionated irradiation in animal been recently reviewed [41e43]; a comparison with animal models [58], where it was shown to eliminate acute hypoxia studies has been made [44]. Although the ‘literature . is (intermittent closure of blood vessels) [59]. The possibility overwhelmingly of its importance as a prognostic factor’ of repair inhibition in normal tissues in vivo does not seem to [36], both the most recent [43] and an earlier discussion be a problem [60], and although it was suggested that small [36] noted the difficulty in defining optimal haemoglobin dose reductions might be needed because of some normal levels. Transfusion with red blood cells often seems to tissue sensitization [61], several promising clinical studies be limited to cases of severe anaemia [43]; it may be have been reported. These mainly involve the combination detrimental in some instances because of immunosuppres- of nicotinamide with normobaric carbogen (to overcome sion [42,43]. A paper [45] discussed the interrelationships diffusion-limited or chronic hypoxia), often with acceler- of low haemoglobin levels, hypoxia, tumour angiogenesis ated radiotherapy to inhibit repopulation (the ‘ARCON’ and survival. Derivatising haemoglobin with polyethylene regimen: accelerated radiotherapy with carbogen and glycol to improve the biocompatibility of bovine haemo- nicotinamide) [36,62,63]. A phase II study of bladder globin was found to be useful in animal models involving cancer found no difference in bowel and bladder morbidity fractionated radiation, particularly in carbogen-breathing at 12 weeks, although there were problems with patients 402 CLINICAL ONCOLOGY completing daily nicotinamide at 80 mg/kg over 20 frac- in desmethylmisonidazole (Ro 05-9963) and etanidazole* tions/4 weeks [64]; another study discussed possible re- (SR2508) were not as successful as hoped [9,82]. Pimonida- lationships between tolerance and pharmacokinetics [65], zole* (Ro 03-8799 [83]) showed enhanced efficacy in vitro in and daily doses of 60 mg/kg became normal. Acute and late (extracellular media) concentration terms compared with morbidity in the treatment of advanced bladder carcinoma misonidazole without correspondingly enhanced cytotoxicity with the ARCON regimen was assessed again more recently [83], but this may have reflected pH gradients between [66], concluding that ‘Although, for some endpoints, the culture media and intracellular milieu [84], or between incidence of late sequelae was higher than expected, overall intracellular organelles [85] (the compound is a fairly strong morbidity was no worse than reported by others. The data base with pKa w 8.7 compared with nimorazole’s pKa w 5.0). indicated that ARCON could achieve a therapeutic gain in Despite good tumour uptake [86], pimonidazole, too, was not patients with advanced bladder carcinoma’. Reviewing found to be effective in trials [36], but it was ‘reborn’ as ARCON in 2002, Kaanders et al. [67] considered ‘In particular a useful diagnostic probe for hypoxia [87]. in cancers of the head and neck and bladder, the local However, drugs were revisited that had been ‘passed by’ tumour-control rates are higher than in other studies .’; because of, with hindsight, over-optimistic expectation of another review [47] confirmed this conclusion, although in realising high dose-modifying effects. There is evidence from unselected groups of patients in a phase I/II study there was the redox dependence of radiosensitization efficiency with no significant difference in tumour response and local electron affinity, which varies with the level of response control with carbogen and nicotinamide added to conven- analysed [80], that there is more than one mechanism of tional radiotherapy [68]. ARCON treatment reduced the radiosensitization. Compounds with low, but still significant, prognostic significance of haemoglobin in squamous cell efficacy in vivo could be tolerated to very high doses. In carcinoma of the head and neck [69]. The outcome of phase particular, the 5-nitroimidazole, nimorazole* was shown to III trials is awaited with interest. be effective in several clinical trials [37,79], and it has been in routine use in the treatment of head and neck cancer in Denmark. It is most unfortunate that its ‘orphan drug’ status ‘Electron-affinic’ Radiosensitizers: the seems to inhibit its wider use despite (or because of?) the cheapness of the treatment. It is clear that the better Nitroimidazoles and Related Drugs selection of patients for inclusion in trials involving treat- In the 1960s, comparison of the chemical properties of ments aimed at hypoxic cells would be beneficial. In the chemicals that radiosensitized anoxic cells with their context of nimorazole, for example, high concentrations of reactivity towards radiation-produced free radicals led to osteopontin in plasma were related to the response seen with the identification of an important class of ‘electron-affinic’ nimorazole in the Danish DAHANCA 5 trial [88]. radiosensitizers [15,16,70]. Early prototypes were nitro- Numerous alternative nitroaromatic structures were benzenes [71,72], but practical drugs based on nitro- evaluated as alternatives to misonidazole and its nitro- imidazoles, such as metronidazole*, already in clinical use imidazole analogues [89], of which the nitrotriazole, as trichomonacidal agents, soon emerged [73]. The best sanazole* (AK-2123), is the most enduring [90]. A large known of these as radiosensitizers is misonidazole* (Ro 07- (462 patients) phase III clinical trial of sanazole in the 0582), which had activity in all solid murine tumour models treatment of squamous cell carcinoma of the uterine cervix tested (to the author’s knowledge): a compilation in 1981 showed an increase in local tumour control and survival listed almost 50 studies [74], involving end points such as without the addition of any major toxicity [91]. (A criticism local control, regrowth delay and cell survival. No sensiti- of many of the early nitroimidazole trials was that they zation of normal tissues was generally found, but mild were too small.) An interesting development of nitro- hypoxia (i.e. sensitization) in skin [75] and effects in mouse aromatic chemistry involving sanazole is its evaluation as tail tissue were shown [76]. It was expected [16], and found a possible radiopharmaceutical agent designed for therapy [74,77], that efficacy was reduced (but not eliminated) in rather than imaging: a derivative of sanazole linked to multi-fractionated treatments because of re-oxygenation a chelating agent complexed with 177Lu has been described between treatments. Cells at intermediate oxygen tensions [92]. (Nitroaromatic compounds bind selectively to hypoxic are especially critical in defining the effects of hypoxic cell cells via complex reductive chemistry involving an in- radiosensitizers [78]. termediate,oxygen-sensitiveradical: numerousstudies aimed However, in the event, toxicity in humans was dose at diagnosing hypoxia and involving immunohistochemical limiting: the efficacy of misonidazole as a radiosensitizer [87,93,94] or radiopharmaceutical [95] detection have was not proven after numerous trials [36], although a meta- been carried out with nitroimidazoles.) Doranidazole (PR- analysis involving trials including over 7000 patients indicated 350) is a nitroimidazole very similar to misonidazole but significant benefit of locoregional control after radiotherapy with greater hydrophilicity; a phase III trial, involving given with nitroimidazole radiosensitizers [79]. Toxicity in only 48 patients, of doranidazole combined with intra- vitro paralleled radiosensitization efficacy, because the same operative radiotherapy in advanced pancreatic cancer, property d electron affinity or reduction potential d dom- did not show significant gain [96]. inated the structureeactivity relationship [80,81]. Attempts The nitroimidazoles and related compounds are often to improve the therapeutic ratio in vivo by increasing termed ‘oxygen-mimetic’ radiosensitizers. There are some hydrophilicity (to decrease peripheral neuropathy), as parallels in their reactivity towards DNA base radicals that CHEMICAL RADIOSENSITIZERS IN RADIOTHERAPY 403 might justify this description. Fig. 4 shows possible [101]. Examples include nitroimidazole/intercalator conju- mechanisms by which both oxygen and nitroimidazoles gates [102,103], nitroacridines (e.g. nitracrine*) and nitro- and related compounds might enhance DNA strand breaks, quinoline intercalators [104], minor groove binders [105], based on many radiationechemical studies [97e100].A and polyamine conjugates [106]. Even pimonidazole was key, but now often overlooked, set of experiments was shown to be ‘concentrated’ near DNA relative to misonida- carried out in the 1970s, which showed that both types of zole [107,108]. However, diffusion-limited hypoxic cells are radiosensitizer had to be present at the instant of distant from capillaries, so free diffusion unimpeded by irradiation: adding either a few milliseconds after irradia- binding to cells nearest capillaries is a desirable property, tion was ineffective [13,14]. This supports a free radical and this targeting approach does not appear to have been mechanism and should steer investigators away from more useful clinically to date. On the issue of diffusion of drug ‘modern’ explanations involving effects of hypoxia on cell from the vasculature to distant hypoxic cells, it is notable signalling pathways, which seem unlikely to be able to be that significant advances in understanding the processes ‘switched’ on or off on this timescale. An important influencing drug delivery and distribution in tumours have property of the ‘electron-affinic’ class is that they appear been made as a result of interest in this issue with to radiosensitize hypoxic cells but have no measurable radiosensitizers and hypoxic cell toxins [109e112]. This effect in well-oxygenated cells, at least in vitro (any small experience is highly relevant to cancer treatment by drugs effects on normal tissues [75] probably reflect tissue in its widest sense. oxygen levels much lower than those commonly encoun- tered in in vitro models). This is probably not because of some remarkable inherent property, but because of simple Aromatic N-oxides: Alternatives to kinetic competition between oxygen and nitroimidazole (or Nitroaromatic Compounds with Most analogue) for reaction with key DNA base radicals. The competition is characterised by the product of reactivity Value as Hypoxic-specific Cytotoxins (chemical rate constant) and concentration: the high Some aromatic N-oxides, especially the benzotriazine di-N- reactivity of oxygen ensures it out-competes the radio- oxide, tirapazamine* (SR4233), have attracted considerable sensitizer. This begs the question as to the probable attention, including several clinical trials, because of competition if drugs are designed to bind to DNA, e.g. by selective cytotoxicity towards hypoxic cells in the absence intercalation or electrostatic interaction. The topic of DNA of radiation [24,113,114]. This is a property also shared targeted radiosensitizing drugs has attracted attention with nitro compounds d metronidazole is prototypical

O O O O O O O O NH • + 2 NH2 NH2 NH2 H • H H H OH H OH H OH strand OH N OH N OH N OH N H H break H - O O N O O N O O• O O N OOH O O N OOH H H H P O P O P O P O O O O O

1 H H 3 H H 4 H H 5 H H OH OH OH OH O P O P O P O P

O2 •OH

O O O O O O O O NH NH • NH + NH 2 2 2 • 2 H H OH H H H H OH OH OH strand OH N OH N OH N OH N H H O H OH break H OH • • - O O N H O O N O N O O N O N O O N O N H H H Ar P O Ar P O Ar P O Ar P O O NO2 O O O

2 H H 6 H H 7 H H 8 H H OH OH OH OH O P O P O P O P Fig. 4 e Possible pathways by which hydroxyl radicals (OH) can add to the 5,6-double bond of pyrimidines (1) to form a carbon-centred radical (2) that can either add oxygen to form a peroxyl radical (3) or add to the oxygen atom in the nitro moiety of a nitroaromatic radiosensitizer

(ArNO2) to form a radical adduct (6). In either case the intermediate radical (3) or (6) might abstract hydrogen from a neighbouring sugar CeH bond (50 in this example, although 30-abstraction may occur) to transfer radical damage from base to sugar (4) or (7), leading to a strand break (5) or (8). Based mainly on a scheme by Bamatraf et al. [100] and extensive studies by other workers described by von Sonntag [97,98]. Only part of the overall mechanism of strand break formation is shown; for further details see [98]. 404 CLINICAL ONCOLOGY

[115], if inefficient d but to a less marked degree. Both alkylating function. Clearly, this will also occur to some types of compound rely for their selectivity as hypoxia- extent even if, for example the drug adds to radiation- specific cytotoxins on the fast reaction between drug produced DNA base radicals to form a radical adduct (cf. radicals and oxygen [116,117]. However, whether benzo- Fig. 4) (the pKa of the piperidine nitrogen in pimonidazole triazine-N-oxides, for example, can also act as hypoxic cell increases from w8.7 to w9.2 even in the radical anion radiosensitizers in a similar manner to oxygen or nitro- [120]). Hence, the possibility of radiation-induced cross- imidazoles (either phenomenologically or mechanistically) linking of DNA as a mechanism needs to be considered, as is less certain, although it seems probable. The only well as ‘simple’ potentiation by enzyme-catalysed re- unambiguous method to answer this point would be to duction. A recent pre-clinical study [128] compared the conduct rapid-mix experiments where both pre- and post- efficacy of tirapazamine and RB6145 in reducing metastatic irradiation exposure are so short that any effects arising dissemination after treatment of the primary tumour with from hypoxic metabolism to reactive radicals can be radiation and drug in a fractionated regimen, providing discounted: oxygen and nitroimidazoles do act on a very evidence that targeting hypoxic cells in primary tumours is short exposure timescale [13,14]. There is absolutely no a valuable strategy to reduce disseminated disease. question that tirapazamine potentiates cell killing with fractionated irradiation, but activity can be measured if Nitric Oxide: the Born-again Radiosensitizer tirapazamine is given after radiation and the data ‘can be largely accounted for by complementary cytotoxicity of the Nitric oxide is included here as ‘oxygen mimetic’ as it two agents’ (drug and radiation) [118]. Notwithstanding the shares with oxygen its properties of being a free radical in fact the chemical mechanisms of radiosensitization by the ‘stable’ form and highly reactive towards many other nitroaromatics, let alone N-oxides, are not well estab- free radicals. However, the similarity ends there. Although lished, if sensitizer efficiency for the N-oxides followed the it reacts rapidly with hydrated electrons produced when same relationship between electron affinity as does nitro- water is irradiated (along with many other chemicals) aromatics [80], then tirapazamine would be intermediate in [129], it has a low electron affinity on the same scale of efficiency between misonidazole and metronidazole, as its measuring oxygen or the nitroimidazoles (reduction poten- reduction potential is close to midway between these tial e0.55 V at pH 7 [130], lower than that of metronidazole nitroimidazoles [119]. Certainly another chemical property [e0.50 V]). More importantly, the unpaired electron is able that might be suggestive d the reactivity of the drug to pair up with that in, for example DNA base radicals, radical anions towards oxygen d puts the aromatic N- forming, unlike DNA base peroxyl radicals from oxygen, oxides firmly on the same redox ‘line’ as nitroaromatics a non-radical species. This is unable, for example, to [120]. As such, considerably higher concentrations of perform the hydrogen abstraction from a sugar, which can tirapazamine might be needed to show significant ‘conven- lead to strand breaks as shown for oxygen and nitro- tional’ radiosensitization than used in most experiments, imidazoles in Fig. 4. A stable adduct can be separated by because of the high potency as a hypoxic cytotoxin. Aerobic high performance liquid chromatography in irradiated radiosensitization after hypoxic pre-incubation is clearly solutions of uracil and nitric oxide [35]. In spite of this linked to drug metabolic activation [121,122]. Descriptions key difference, nitric oxide enhances the yields of DNA of related N-oxides as ‘hypoxic cell radiosensitizers’ are double-strand breaks in irradiated cells [35], as do oxygen being made [123], but should be viewed in the context of and nitroimidazoles [131] (but note: there is not always these difficulties in interpretation. These comments should correlation between double-strand breaks and clonogenic not detract from the potential value of N-oxides as hypoxic survival [132]). The mechanism in the case of nitric oxide is cytotoxins, which are obvious and of clinical importance. not at all clear, but might involve, for example, excision of the damaged base (which could perhaps be proven using cells deficient in base excision repair proteins); the repair Radiosensitizers that are both times in the popular g-H2AX assay seemed longer for cells ‘Electron-affinic’ and Efficient irradiated in nitric oxide compared with cells irradiated in air or anoxia [35], although further work is needed to verify Hypoxia-specific Cytotoxins this. ‘Dual-function’ radiosensitizers, where the drugs can act The most remarkable property of nitric oxide as a hypoxic like misonidazole in modifying free radical damage, and cell radiosensitizer is its efficiency. It was shown to be an also include a second cytotoxic functionality such as active radiosensitizer 50 years ago [133,134], but resur- alkylating or binding, activated metabolically (and possibly rected by Mitchell et al. [135e137] in timely work after the radiolytically), are another class of agent where the basis discovery of its other remarkable property in the mainte- for activity is not always easy to separate. The best known nance of vascular tone [138]. Recent studies by the author’s of these are the 2-nitroimidazoles with an aziridine moiety colleagues, illustrated in Fig. 3, have shown the effects on on the sidechain, such as RSU1069*, or the prodrug for this cell survival in vitro using clinically relevant radiation compound activated hydrolytically, RB6145* (CI-1010) doses, of only 40 ppm v/v nitric oxide (w70 nM), similar to [124e127]. Enzyme-catalysed reduction of the nitro group that used in inhaled nitric oxide therapy for respiratory in hypoxic cells positively shifts the pKa of the aziridine conditions [35]. In this study, at low radiation doses, nitric nitrogen in the sidechain of RSU1069 to activate the oxide seemed to be significantly more efficient than oxygen CHEMICAL RADIOSENSITIZERS IN RADIOTHERAPY 405 as a hypoxic cell radiosensitizer. The rapid removal of nitric activation of the interferon-g pathway and induction of oxide by reaction with red blood cells maintains a low nitric oxide synthase [158]. steady-state concentration in vivo. However, the potential for radiosensitization by nitric oxide by elevating its synthesis or delivery in vivo has been shown in a series Radiation-induced Delivery of Cytotoxic of studies involving gene therapy [139e143], and others Substances using a chemical source of nitric oxide [144] or physical stimulus [145]. This approach to radiosensitization exploits the increasingly Measurements of nitric oxide levels in human tumours sophisticated targeting of radiation delivery in a most are lacking d clinical measurements of tumour oxygena- direct manner, using radiationechemical reactions to de- tion became available only decades after the discovery of liver cytotoxins to the irradiated volume. It has received its importance d but one study using a microelectrode in limited attention. One problem is that the bulk of the the B16 melanoma model [146] suggested levels much radiation energy is absorbed in the bulk of the material higher than shown to enhance anoxic cell radiosensitivity comprising the cell (water), generating initially less than in vitro in our recent work. This raises two hypotheses 0.3 mM reducing radicals per gray dose (mainly hydrated worthy of addressing: first, variations between individual electrons) and the same amount of oxidizing radicals patients in levels of nitric oxide in tumours have clinical, (mainly hydroxyl radicals). Both species are promiscuous prognostic significance in radiotherapy; second, giving in their reactivity and it is difficult to intercept their nitric oxide by inhalation could be useful in the radiother- reactions, except by, often, unrealistically high concentra- apy of lung tumours. tions of drugs (‘scavengers’). Hence, if, for example, the There could be an immediate obstacle to developing approach is to use radiation to liberate a cytotoxin from nitric oxide delivery, or in situ generation by either giving a prodrug in the same manner as the ‘triggereeffector’ prodrugs for nitric oxide or enhancing nitric oxide synthase concept successfully applied in hypoxia-selective, enzyme- using gene therapy or other protocols. This is the existence activated prodrug fragmentation (see above), then the of the ‘steal effect’, by which systemic blood pressure cytotoxin must be very potent indeed. Nonetheless, there reduction can make the tumour more hypoxic, reported in could be some selectivity to hypoxic cells if reductive studies of the effects of a nitric oxide prodrug in rat activation is considered, via electrons directly or reducing tumours [147], and discussed in a recent review of nitric radicals formed on reaction with hydroxyl radicals with oxide in the context of tumour biology [148]. However, amino acids (for example), as oxygen reacts rapidly with there are two counter-claims. First, in the case of inhaled both [97,98]. As with ‘electron-affinic’ radiosensitizers, nitric oxide, there are reports of the selectivity of redox properties are critical as superoxide radicals can be vasodilation to the lung [149]; systemic vascular resistance reducing. was not reduced in mice [150]. Second, nitric oxide One group has carried out quite extensive studies, mainly enhancement by gene therapy actually works in a murine in vitro, involving two types of redox switch. The first tumour model of fractionated radiotherapy [143]. approach seeks to exploit the well-known rapid fragmen- Notwithstanding the ‘steal effect’, one might have tation of nitroaromatic radical anions substituted with considered nitric oxide therapy as a route to improve oxygen ‘leaving groups’, the simplest of which is halogen [159], the delivery for radiosensitization, via effects on vascular tone. practical examples in this instance being a quaternary Paradoxically, a recent clinical study [151] showed that the nitrogen centre in conjugates of nitroaromatics and inhibition of nitric oxide synthase by pharmacological nitrogen mustards [160]. The chemistry proved rather intervention resulted in a reduction in tumour blood volume complex, with great sensitivity to the nitroarene ‘trigger’ that could have therapeutic significance: there is much [161]. The second approach explores the potential of current interest in targeting the tumour vasculature cobalt(III) as the redox target for radiation-produced [152,153]. Whether such approaches have their main role reducing radicals, exploiting the known rapid fragmenta- in chemotherapy rather than radiotherapy remains to be tion of cobalt(II) complexes [162,163]. In the latter study, seen; Denekamp [154] commented on the limited role of the drug released on radiolytic activation was a potent DNA vascular-mediated injury in tumour response to radiother- minor groove alkylating agent, a ring-opened analogue of apy, but this pre-dates recent observations of substantial CC-1065. Although both approaches are well-founded and selective damage to tumour vasculature with some mechanistically, exemplification in vivo is crucial. treatments. There are several possible effects of nitric Another group evaluated the application of radical oxide on treatment and prognosis in cancer fragmentation after capturing radiation-produced reducing [148,151,155,156], and this is an area meriting further equivalents, using conjugates of indolequinones and 5- study. Effects can be indirect; thus, radiosensitization by fluoro-20-deoxyuridine [164], or the latter bearing a 2-oxo- insulin treatment was ascribed to an increase in tumour alkyl substituent as an electron-attracting centre [165]. pO2, not caused by increased tumour blood flow but Significant effects of the latter approach were not shown ascribed to a decrease in oxygen consumption caused by in vivo using a regrowth delay assay. The very low electron nitric oxide [157]. OM-174, a synthetic analogue of Lipid affinity of an oxo-alkyl centre (cf. acetone or acetophe- A, the endotoxic principle of lipopolysaccharides, was none) means that the latter type of prodrug has to rely on shown to radiosensitize EMT-6 tumour cells in vitro via capturing electrons for activation; whereas all the other 406 CLINICAL ONCOLOGY studies also seem to focus on (hydrated) electron capture, have been investigated as radiosensitizers (including this emphasis is probably misplaced, as a wide variety of clinical trials), although the basis for these is controversial reducing radicals formed via oxidative damage can reduce [174,175]. The lead compound is motexafin gadolinium nitroarenes, cobalt(III) complexes, or indolequinones. (PCI-0120). The complexes are reactive towards reducing Other redox metals have been considered as radio- radicals, electrochemical studies revealing a reduction sensitizers. Copper(II) complexes were first studied many potential of we0.08 V on the same scale as other radio- years ago [166,167], and interest recently resumed [168]. sensitizers [174,176]. This value seems on the high side for The use of such redox-active complexes is paved with redox cycling (reducing oxygen to superoxide/hydrogen difficulties to investigate in vitro because of the well- peroxide) [177], as the corresponding reduction potential known artefacts that can arise from reactions of the metal for oxygen (1 M) to superoxide is e0.18 V [178,179]. In any with hydrogen peroxide produced in the irradiated media, case, the suggestion [177,180] that sensitization to radia- or with thiols in the media as well as in the cells, and the tion by motexafin gadolinium arises from increasing diverse roles of copper in redox biology [169,170]. Although oxidative stress via redox cycling should be viewed with copper complexes are feasible drugs [171], until there is caution. Other ‘electron-affinic’ radiosensitizers redox clear in vivo exemplification, this approach seems a long cycle rather efficiently, but the process is clearly not way from clinical interest. responsible for their efficacy; more probably, some A problem generally in this area is that most in vitro pathological responses are linked to the phenomenon models involve irradiating cells in a large volume excess of [21]. It was noted elsewhere that cells normally generate medium, usually containing drug unless intracellular: as much oxidative stress (superoxide) in about 20 s as is extracellular concentration gradients are very high. produced by 1 Gy irradiation [181] d although, of course, Radiationechemical events in the medium may produce mitochondrial production is highly compartmentalized, a large excess of released drug that can diffuse into cells whereas radiolytic generation is not. The reaction of and amplify damage, a scenario that cannot be extended to motexafin gadolinium with cellular reductants, including anything like the same extent in vivo. This lesson is also ascorbate and glutathione [177], which will be accompa- pertinent to radiosensitizer (and general drug) research in nied by superoxide formation (copper(II) behaves simi- a wider sense. Early experience with thiol-reactive nitro- larly), should be viewed in the context of the potential imidazoles [172] showed the artefacts in extrapolation that difficulties noted above in extrapolation of phenomena can arise when a huge molar excess of reagent compared seen in vitro. Ascorbate was shown to be an important with target (e.g. cellular thiols) is involved in in vitro enhancer of effects in vitro [182], including DNA damage models. Indeed, this author suggested that neglect of this [183]. The absence of ascorbate might explain in part the factor d which he termed ‘simple arithmetic’ as it can be negative results in some studies [175], but some effect on done on the back of a postage stamp, let alone an envelope tumour regrowth delay with combined gadolinium/radia- [173] d may be responsible for many false leads in drug tion treatment has been reported [184]. It is not clear research. The principle of this arithmetic test for the whether this effect arises from ‘direct’ radiosensitization potential lack of translation of effects in vitro to reality in or enhancement in tumour oxygenation [185]. A recent vivo is illustrated in Table 1. Basically, adding very low review suggests ‘the molecular target . appears to be concentrations of (in this example) a thiol-reactive chem- thioredoxin reductase’ [186]. Clinical trials with motexafin ical to cells in dishes wipes out all the protective thiols, gadolinium have been summarized [54,186,187], most while still leaving plenty of drug to radiosensitize, but the involving the treatment of brain metastases [188]. In phase effect will never be realisable in vivo because the III studies, an improved time to neurological progression in arithmetic does not add up: the moles of drug cannot patients with brain metastases receiving whole brain compete with the moles of endogenous thiol. radiotherapy with gadolinium has been reported. Overall, Gadolinium(III) is a redox metal used as a contrast agent the mechanistic basis for the action of gadolinium in magnetic resonance imaging. Porphyrin-like complexes complexes as radiosensitizers is not entirely clear d it (‘texaphyrins’) of gadolinium(III) localise in tumours, and seems doubtful whether it should indeed be classified under this section as delivering cytotoxic substances d and Table 1 e Thiol depletion in vitro and in vivo after systemic further studies are required to demonstrate the potential administration of 1 mM or up to 3 g, respectively, of a thiol- of this approach. reactive chemical with molecular weight 200 [173] Porphyrins have long been used as photosensitizers in photodynamic therapy [189], but also have radiosensitizing In vitro In vivo (e.g. fibroblasts) (e.g. humans) properties; limited clinical trials with radiotherapy are being conducted [190e192]. The most extensively studied example in photodynamic therapy is probably Photofrin II, Cell density w104e105/cm3 w5 108/cm3 a complex mixture of porphyrins concentrated in, or Free thiols w5 fmol/cell w2e5 mmol/kg Thiol-reactive reagent w1 nmol/cm3 media ! 3g retained by, some tumours relative to some normal tissues. administered (example) w100e10 fmol/cell !0.2 mmol/kg Their efficacy as radiosensitizers, even in vitro, is quite Thiol depleted 100% !10% dependent on the tumour cell line [191], in contrast to Reagent depleted w5e50% 100% radiosensitizers such as misonidazole. The mechanism underlying radiosensitization is completely unknown, but CHEMICAL RADIOSENSITIZERS IN RADIOTHERAPY 407 in view of the established use of these agents in photo- recent review [200] updates these. Discussing possible medicine, further work is certainly justified. mechanisms, McGinn and Lawrence [197] noted that ‘In- corporation of . BrUrd and IdUrd into DNA has been associated with increased induction and decreased rate of Radiosensitization by Halogenated repair of radiation-induced DNA damage. Studies using Pyrimidines similar techniques have found no effect on radiation- induced DNA damage or repair after exposure to gemcitabine This approach can be divided into two distinct categories, under conditions known to produce radiosensitization. . In essentially distinguished by halogen: bromine/iodine vs contrast, data tend to support the hypothesis that gemcita- fluorine. First, cells undergoing DNA synthesis cannot bine-mediated radiosensitization is related to concurrent distinguish efficiently between thymidine and halogenated disruption of deoxyribonucleotide pools and redistribution of analogues such as bromo- or iodo-deoxyuridine (UdR) (the cells into S phase of the cell cycle’. (Cells are usually most methyl group of thymine is about the same size as bromine radiosensitive close to mitosis, and most radioresistant in or iodine atoms). If cells are treated with BrUdR or IUdR late S phase [17].) Gemcitabineeradiation interactions are for a sufficiently long period before irradiation, significant complex: reviewing pre-clinical data and clinical trials, it incorporation into DNA occurs. The halogen moieties act as was noted [201] that ‘the mechanism of radiosensitization by electron ‘sinks’ on irradiation, the carbonehalogen bond gemcitabine is still not fully elucidated, and the optimal breaking on electron attachment to liberate free halide treatment schedule still has to be defined’. The need for and form a carbon-centred free radical. This can add phase II trials of radiation with gemcitabine was stressed [8]. oxygen to form a peroxyl radical and carry out similar A succinct but comprehensive survey of the mechanistic strand-breaking reactions as the DNA base/hydroxyl radical basis for the use of both 5-FU and analogues, and adduct illustrated in Fig. 4. There is a correlation between gemcitabine, stressed the importance of inappropriate BrUdR incorporation, DNA strand breaks and clonogenic progression of cells into S phase [3]. survival after irradiation [193]. Skin phototoxicity is A review of radiosensitizing nucleosides [196] included a problem with BUdR, but not with IUdR. Although success a summary of experience with hydroxyurea as a radiation in phase III clinical trials has been elusive, the potential sensitizer because its primary mechanism of cytotoxicity is application of this approach in (low dose rate) brachy- related to the inhibition of ribonucleotide reductase. A therapy is of particular interest; a review of the laboratory number of clinical trials, with mixed results, were and clinical studies to 2001 summarised both positive and summarized. Hydroxyurea can also modulate both fluoro- negative results [6]. More recent clinical attention has pyrimidine- and IUdR-mediated radiosensitization. focused on the iodinated analogue, and particularly on a prodrug, 5-iodo-2-pyrimidinone-20-deoxyribose, which is converted to IUdR by an aldehyde oxidase in the liver Radiosensitizers that Influence the [194]. It was suggested that this approach is suitable for Nature or Repair of DNA Damage drug-resistant, DNA-mismatch repair-deficient (as well as repair-proficient) tumours; IUdR and BrUdR accumulated Categorising radiosensitizers into well-defined types is at much higher levels in mismatch repair-deficient cells [195]. particularly difficult when some chemicals have dual More recent approaches to the use of radiosensitizing effects. Nitroimidazoles with sidechain alkylating function- nucleosides have focused on fluorine analogues, especially ality (e.g. RSU1069) are a simple example (see above). The 5-fluorouracil (5-FU), 5-FUdR, gemcitabine* (20,20-difluoro- use of redox metals with possible dual functionality 20deoxycytidine), and a rationally designed prodrug of discussed above (copper, gadolinium) is another; platinum 5-FU, capecitabine* [196,197]. The latter exploits the a third. Cisplatin and carboplatin have been studied in differential activity of thymidine phosphorylase in tumour some detail in combined treatments with radiation, with compared with normal tissue in the final of three steps in activity apparently as ‘conventional’ electron-affinic hyp- converting prodrug to active 5-FU; it was suggested [198] oxic cell radiosensitizers and as inhibiting post-irradiation that capecitabine ‘has the potential to replace bolus or repair [202,203]. There was a time-dependent increase in continuous infusion of 5-FU as the standard treatment [in the extent of DNA double-strand breaks after irradiation of chemoradiotherapy] for rectal cancer’ (oral 5-FU has un- several cell lines in vitro with high concentrations of predictable bioavailability [197]). Another paper reviewed carboplatin [204]. Dual-function platinum/nitroimidazole the use of capecitabine and other agents as radiosensitizers complexes have also been evaluated [205], extended to in rectal cancer [199]. The mechanisms of the radiosensitiz- non-platinum analogues [206,207]. A recent detailed ing effects of the fluorinated analogues are clearly com- discussion clearly set out the several mechanisms that pletely different to the bromo- and iodo-nucleosides, might be involved in the interaction of platinum complexes presumably reflecting at least in part the different ‘leav- with radiation [3]. Numerous clinical trials have been ing-group’ abilities of the halogens and van der Waals’ radii reported involving cisplatin or carboplatin and radiother- (atomic size), as well as (of course) differences from apy, often with additional cytotoxins and usually involving substitution of halogen in base or sugar. The October 1997 drug treatment with a few of the radiation fractions issue of Seminars in Radiation Oncology includes 10 papers because of toxicity. An illustrative, recent review of the discussing relevant aspects of fluoropyrimidines; a more chemoradiation paradigm in the context of non-small cell 408 CLINICAL ONCOLOGY lung cancer [208] summarised the position as ‘The use of the specificity and activity in vivo to enhance radiotherapy single-agent cisplatin has already demonstrated major [218]. Treatment with AG14361* before irradiation signifi- radiosensitizing effects whereas the radiosensitizing cantly increased the sensitivity to radiation therapy of mice properties of concurrent application of the single agent bearing LoVo xenografts: irradiation alone (2 Gy daily for carboplatin have not been observed in controlled trials’. 5 days) caused a 19-day tumour growth delay, extended to Another review [209] focused on squamous cell carcinomas a 37-day delay in mice treated with AG14361 before of the head and neck and oesophageal carcinomas, irradiation. concluding with respect to the former site that concomi- The ataxia telangiectasia mutated (ATM) protein kinase tant chemo- and radiotherapy with platinum-based drugs plays a critical role in regulating cell cycle arrest and DNA does provide a small survival gain compared with radiation repair, linked to dramatically enhanced radiation sensitiv- alone, and for the latter site that preoperative, cisplatin ity, and is attracting attention as a target for radio- chemoradiotherapy modestly improves outcome over sensitizers [219]. Inhibitors of the ATM kinase, such as surgery alone, but with additional morbidity. Early trials wortmannin and caffeine, sensitize cells in vitro; the involving radiation with a newer analogue, oxaliplatin, have former is very reactive towards proteins non-specifically, been summarised [7]. including effects on enzymes catalysing non-homologous b-Lapachone is a naturally occurring 1,2-naphthoquinone end-joining (an important DNA repair mechanism in normal and as such might be considered an electron-affinic radio- cells), and the latter is active only at undesirably high sensitizer. However, it was shown to inhibit topoisomerase I concentrations (around 0.2 mM). Pentoxifylline is a drug and radiosensitize human malignant melanoma cells in vitro related to caffeine that has also been shown to have when added after irradiation [210]. Other DNA topoisomer- vasoactive properties and to modify tumour oxygenation as ase I-targeted drugs have been described as radiation described above [56,57]; studies of both caffeine and sensitizers effective when given before, but not after, pentoxifylline in the context of ATM as a target have been radiation, including derivatives of camptothecin* [211]. reviewed [219]; neither appeared clinically useful, al- Although the mechanism of topoisomerase I-mediated though a more recent overview [8] was more optimistic. radiosensitization ‘remains [in 2004] largely unknown’ However, there are some interesting points. Caffeine is [211], a role for Ku86, important in the non-homologous more effective as a radiosensitizer in cells that lack normal end-joining pathway in DNA repair, has since been shown p53 function, by a factor of 1.4e2.8-fold, and another [212]. Topotecan is a derivative of camptothecin, which methyl xanthine, lisofylline*, radiosensitizes cells lacking inhibits topoisomerase I in S phase cells; it has been p53 at clinically tolerable concentrations [220]. suggested that it may be effective in the treatment of brain metastases by radiation [213]. AQ4 is an anthraquinone that is a DNA intercalator and topoisomerase II poison; the Radiosensitizers Targeting Proteins di-N-oxide prodrug, AQ4N*, is activated selectively in Involved in Cell Signalling, and hypoxic cells, and when combined with radiation greatly Growth Factors enhances growth delay in murine tumours [214]. In related work, intratumour injection of an activating cytochrome This family of radiosensitizers is the newest to be CYP2B6 vector showed further enhancement over AQ4 N investigated, although why they are often termed ‘molec- and radiation [215]. The agent is currently in phase ular agents’ or the targets ‘molecular’ is unclear: misoni- I/II clinical trial in combination with radiotherapy and dazole and DNA are molecules too. Three recent reviews temozolomide in patients with glioblastoma multiforme. focusing on this area illustrate the diversity of approaches Temozolomide* is a prodrug for an alkylating agent that [7,221,222]. Proteins involved in cell signalling and growth methylates guanine at O-6; clinical trials of temozolomide factor receptors are the main categories of target. In the plus radiation in malignant glioma have recently been former, the proteins most studied in the context of summarised [216]. In a large phase III European study, the radiation therapy are probably the Ras family, which is of 2-year progression-free survival was increased from w2to wider interest in cancer therapy because they control 11% by the inclusion of temozolomide with radiotherapy, signalling pathways that are key regulators of cell growth and overall survival at 2 years from w10 to 27%, ‘one of and transformation [223]. There is substantial evidence the most significant improvements in survival found in linking Ras proteins to radioresistance [23,224]. The any phase III clinical study in newly diagnosed GBM functionality of Ras is in turn linked to the addition of [glioblastoma multiforme] patients within the past several a farnesyl isoprenoid moiety, and farnesyl transferase decades’ [216]. inhibitors such as L-778,123* have been evaluated as PARP is a nuclear enzyme facilitating DNA base excision radiosensitizers [225]. Pre-clinical studies showed that such repair; the therapeutic potential of PARP inhibitors in inhibitors can reverse radioresistance in human tumour a broad context has been reviewed [217]. Whereas the xenografts expressing the mutant Ras oncogene, and phase paper noted the early work with nicotinamide in this I/II clinical trials recently commenced. In one study, context, emphasising its low activity as a PARP inhibitor and L-778,123 was used with concurrent radiotherapy to treat thus reinforcing the assignment of effects on tumour patients with locally advanced pancreatic cancer. The data oxygenation to other mechanisms (see above), a recent confirmed pre-clinical observations that there was no study described a much more potent PARP inhibitor that has increase in radiation-induced normal tissue damage, and CHEMICAL RADIOSENSITIZERS IN RADIOTHERAPY 409 therefore the approach has the potential to increase the [236e239]. Some agents were both electronic-affinic and therapeutic index. depleted thiols, and attention has been drawn in the above There is intense activity in the area of epidermal growth discussion (and in Table 1) to the artefacts that can arise in factor receptor (EGFR) inhibitors, which might mediate cell testing these chemicals in vitro because reaction binds or growth, differentiation and survival [3,8,226]. The poten- oxidises all the cellular thiols while leaving a large fraction tial mechanisms of radiosensitization by EGFR are complex of reagent unchanged. This gives rise to an elevated [226]. Enhanced anti-tumour activity with vandetanib* expectation of efficacy impossible to achieve in vivo.In (ZD6474) and radiation was reported in a pre-clinical study addition to the nitroimidazoles where thiols displace [227] using a model previously found to be unresponsive to halogen or other ‘leaving groups’ [172], powerful electron- one selective EGFR tyrosine kinase inhibitor. However, withdrawing substituents, such as eCHO, can activate the scheduling of ZD6474 relative to radiotherapy had a pro- nitro substituent in 2-nitroimidazoles to displacement by found effect on the enhancement: sequential, chronic thiol, catalysed by glutathione-S-transferases [240]. administration of ZD6474 after the radiation treatment A much more sophisticated and controllable approach to significantly enhanced growth delay. In contrast, the depleting intracellular thiols is the inhibition of steps in response after concurrent treatment with ZD6474 and their biosynthesis: the application of L-S-buthionine sul- radiotherapy revealed no significant interaction between phoximine* (BSO) was a major advance [241]. A significant the two modalities. A positive phase II trial evaluating an enhancement in the radiosensitizing efficiency of misoni- anti-EGFR antibody, cetuximab, and radiotherapy, involving dazole in vitro was seen on pre-incubation with BSO to 424 patients with advanced head and neck cancer, has been deplete cellular glutathione (GSH) [242]. However, this was reported [228].Atw54 months, the median duration of not generally translated to in vivo models to anything like overall survival was 49 months for combined therapy and the same extent, despite protocols showing good depletion w29 months for radiotherapy alone, but an editorial of average tumour GSH levels [243e246]. An explanation expressed caveats [229]. Harari and Huang [230] reviewed might have been the poor diffusion of the rather polar BSO the use of cetuximab and other EGFR inhibitors, including molecule to hypoxic cells distant from the vasculature; this gefitinib* (Iressa), noting the highly promising pre-clinical seems not to be the case, although heterogeneity of GSH data, but concluding that ‘the overall clinical gains to date levels in tumours was indicated [247]. Microregional . are modest for the global cancer population’. heterogeneity of GSH in cervical carcinoma, with higher Cyclooxygenase (COX)-2 is often over-expressed in GSH levels more susceptible to BSO-initiated depletion in cancer and is linked to resistance to cytotoxic agents. hypoxic areas, has been reported [248,249]. The inhibition of COX-2 has been found to enhance the BSO has been evaluated in several clinical trials, mostly tumour response to radiation in pre-clinical studies, and in the context of chemotherapy regimens where GSH is COX-2 expression has been linked to tumour radioresistance protective, for example with melphalan [250,251], which [8]. An example being evaluated clinically in the radiother- have shown the clinical feasibility of depleting tumour GSH apy of lung cancer is celecoxib* [231], which was shown in levels. There do not seem to have been recent studies recent mechanistic studies [232] to down-regulate the involving BSO and radiotherapy, although there remains expression of Ku70 protein and to inhibit the kinase activity active interest in alternative approaches to deplete GSH in of DNA-PKcs, important in repairing double-strand DNA tissues [252], and new strategies to confer tumour breaks. It was also suggested that NFkB may play a role in selectivity would have obvious application in both radio- mediating the effects of celecoxib [232]. therapy and chemotherapy. It should be born in mind that Finally, other targets that are attracting attention in the GSH is not the only cellular thiol: levels of both cysteine context of radiation therapy include anti-angiogenesis [253,254] and protein thiols [255] must also be considered inhibitors [233]; a sulphoglycolipid has been identified as in this context. GSH is not the only antioxidant, either: a candidate radiosensitizer [234]. The molecular target is ascorbate is highly reactive towards oxidizing DNA base unknown, although the agents are inhibitors of DNA poly- radicals [256], and competitive radiation dose modification merases. The potential use of inhibitors of the hypoxia- by ascorbate and misonidazole in thiol-depleted cells has inducible factor HIF-1 has been discussed [54]. Vascular been reported [257]. endothelial growth factor and Rad51 enzymes (catalysing DNA double-strand break repair) are other new targets for radiosensitization, discussed in a recent review [200]. Conclusions Membrane targeted drugs have been reviewed as putative radiosensitizers [235]. There are a wide variety of routes by which chemicals can interact in some way with radiation damage to offer potential therapeutic gain. The levels of early chemical Suppression of Radioprotective damage such as DNA strand breaks can be enhanced, as Substances with oxygen, ‘electron-affinic’ compounds and nitric oxide. The tumour microenvironment can be modified to reduce Depletion of intracellular thiols is the obvious approach to acute hypoxia, as with carbogen and nicotinamide. Radio- take in this context [9]. Indeed, thiol-reactive chemicals resistant subpopulations of cells can be targeted, such as were among the first radiosensitizers to be studied hypoxic cells with tirapazamine. The efficiency of DNA 410 CLINICAL ONCOLOGY repair can be inhibited, with the potential to exploit In conclusion, there are several new approaches in the hypoxia to deliver the new generation of potent repair field of chemical radiosensitizers that show promise, but inhibitors (or other ‘molecularly targeted’ drug) selectively their mechanistic basis is poorly researched. Some almost- to tumours via the ‘triggereeffector’ concept, or as forgotten radiosensitizing chemicals, such as nitric oxide, directly hypoxia-activated prodrugs. Tumour proliferation are presenting exciting new possibilities, and there is still can be exploited by S phase-specific uptake of halopyr- life left in the ‘old dog’, oxygen. However, it is no use imidines, which can act as dissociative electron ‘sinks’ to focusing on 21st-century cell biology while neglecting even enhance radical damage when incorporated into DNA or 1950s’ chemistry, or indeed seeking to test sophisticated affect nucleoside and nucleotide metabolism. Enzymes chemistry using inappropriate biological models. involved in signal transduction pathways, cell cycle . checkpoints, growth factors, etc. can all be targeted. Acknowledgements. This work was supported by Cancer Re- To change the status of radiosensitizing chemicals into search UK Programme Grant C134/A7428. The author gratefully clinically useful drugs raises many questions. Coleman and acknowledges the help and stimulus of all his present and former Mitchell [258] summarised these (selecting and paraphras- colleagues at the Gray Laboratory and Mount Vernon Hospital. ing some key points or illustrations in brackets) as: Under the leadership in the laboratory of Ged Adams, Juliana Denekamp, Jack Fowler, and Barry Michael, and in the clinic of ‘What is the target of the modifier? [DNA, transcription Stanley Dische, Peter Hoskin, and Michele Saunders, chemical factor, enzymes, receptor, signalling molecule . ] radiosensitizers were translated from chemical curiosity into Is the target stable? [cell cycle variation, heterogene- radiotherapeutic reality with extraordinary rapidity. This environ- ment of seamless multidisciplinary science in a clinical setting, and ity, resistance . ] these individuals, defined what translational research should be Can the target be reached? [pharmacology, distribution like; it was a privilege to play a small part. (macro/microscopic) . ] What is the optimal schedule? [pharmacokinetics, cell Author for correspondence: P. Wardman, University of Oxford, . cycle perturbation ] Gray Cancer Institute, PO Box 100, Mount Vernon Hospital, Can the radiation modifier be used throughout a course Northwood, Middlesex HA6 2JR, UK. E-mail: [email protected] of fractionated radiation therapy? [drug toxicity, every treatment or selected; if latter, when? . ] Received 27 February 2007; accepted 13 March 2007 Selectivity: tumor versus normal tissue? [is modifier plus radiation better than more radiation . ] What is the design of the clinical trial?’ [choice of end References point, sufficiently large study . ]. 1 Steel GG. Terminology in the description of drug-radiation e It is thus obvious that many skills have to be brought into interactions. Int J Radiat Oncol Biol Phys 1979;5:1145 1150. 2 Steel GG. Basic clinical radiobiology. London: Arnold 1997. play to be successful in this area. Reviewing in 1996 the 3 Wilson GD, Bentzen SM, Harari PM. Biologic basis for combining treatment of cancer with radiation and drugs, Tannock [5] drugs with radiation. Semin Radiat Oncol 2006;16:2e9. concluded: ‘Clinical gains from combined treatment with 4 Seiwert TY, Salama JK, Vokes EE. The concurrent chemo- radiation and drugs have been small. New, mechanistically radiation paradigm d general principles. Nat Clin Pract Oncol based approaches to combined treatment are required’. In 2007;4:86e100. preparing the present overview, the author has been 5 Tannock IF. Treatment of cancer with radiation and drugs. reminded that there has been little advance in the last J Clin Oncol 1996;14:3156e3174. two decades in understanding the mechanisms of how the 6 Poggi MM, Coleman CN, Mitchell JB. Sensitizers and protectors even well-established chemical radiosensitizers actually of radiation and chemotherapy. Curr Probl Cancer 2001;25: e work at the molecular level. Some newer radiosensitizers 331 411. 7 Spalding AC, Lawrence TS. New and emerging radiosensitizers have reached clinical trials without the mechanistic basis and protectors. Cancer Invest 2006;24:444e456. for their putative action being properly understood. 8 Horsman MR, Bohm L, Margison GP, et al. Tumor radio- The limitations of in vitro models should not be sensitizers d current status of development of various overlooked. The author is particularly concerned that approaches: report of an International Atomic Energy Agency ascorbate is present at high concentrations in mammalian meeting. Int J Radiat Oncol Biol Phys 2006;64:551e561. tissues [259], yet normally completely absent in in vitro 9 Coleman CN, Turrisi AT. Radiation and chemotherapy sensi- models used in radiobiology. Ascorbate is well known to tizers and protectors. Crit Rev Oncol Hematol 1990;10: radiation chemists as highly reactive towards some DNA 225e252. radicals [256], as an important ‘radical sink’ in biology 10 Grdina DJ, Murley JS, Kataoka Y. Radioprotectants: current e more generally [260], and as a potential pro-oxidant with status and new directions. Oncology 2002;63(Suppl. 2):2 10. 11 Weiss JF, Landauer MR. Radioprotection by antioxidants. Ann redox metals [169]; as noted above, adding ascorbate NY Acad Sci 2000;899:44e60. diminishes radiosensitization by misonidazole in thiol- 12 Thomas CT, Ammar A, Farrell JJ, Elsaleh H. Radiation depleted cells in vitro [257]. As attention focuses on modifiers: treatment overview and future investigations. newer targets in cell biology, it is important not to Hematol Oncol Clin North Am 2006;20:119e139. overlook simpler molecules such as ascorbate and nitric 13 Shenoy MA, Asquith JC, Adams GE, Michael BD, Watts ME. oxide. Time-resolved oxygen effects in irradiated bacteria and CHEMICAL RADIOSENSITIZERS IN RADIOTHERAPY 411

mammalian cells: a rapid-mix study. Radiat Res 1975;62: 36 Saunders M, Dische S. Clinical results of hypoxic cell radio- 498e512. sensitization from hyperbaric oxygen to accelerated radio- 14 Michael BD, Harrop HA, Maughan RL. Fast response methods in therapy, carbogen and nicotinamide. Br J Cancer 1996; the radiation chemistry of lethal damage in intact cells. In: 74(Suppl. XXVII):S271eS278. Okada S, Imamura M, Terashima T, Yamaguchi H, editors. 37 Overgaard J, Horsman MR. Modification of hypoxia-induced Radiation research (Proceedings of the Sixth International radioresistance in tumors by the use of oxygen and sensitizers. Congress), Tokyo, Japanese Association of Radiation Research Semin Radiat Oncol 1996;6:10e21. 1979;:288e296 38 Feldmeier JJ. Hyperbaric oxygen for delayed radiation in- 15 Adams GE. Chemical radiosensitization of hypoxic cells. Br juries. Undersea Hyperb Med 2004;31:133e145. Med Bull 1973;29:48e53. 39 Koch CJ, Oprysko PR, Shuman AL, Jenkins WT, Brandt G, 16 Fowler JF, Adams GE, Denekamp J. Radiosensitization of hypoxic Evans SM. Radiosensitization of hypoxic tumor cells by cells in solid tumours. Cancer Treat Rev 1976;3:227e256. dodecafluoropentane: a gas-phase perfluorochemical emul- 17 Hall EJ. Radiobiology for the radiologist. Philadelphia: sion. Cancer Res 2002;62:3626e3629. Lippincott Williams & Wilkins, 2000. 40 Chang TMS. Evolution of artificial cells using nanobiotechnol- 18 Denny WA, Wilson WR, Hay MP. Recent developments in the ogy of hemoglobin based RBC blood substitute as an example. design of bioreductive drugs. Br J Cancer 1996;74(Suppl. Artif Cells Blood Substit Immobil Biotechnol 2006;34:551e556. XXVII):S32eS38. 41 Harrison L, Blackwell K. Hypoxia and anemia: factors in 19 Naylor MA, Thomson P. Recent advances in bioreductive drug decreased sensitivity to radiation therapy and chemotherapy? targeting. Mini Rev Med Chem 2001;1:17e29. Oncologist 2004;9(Suppl. 5):31e40. 20 Jaffar M, Abou-Zeid N, Bai L, et al. Quinone bioreductive 42 Varlotto J, Stevenson MA. Anemia, tumor hypoxia, and the prodrugs as delivery agents. Curr Drug Deliv 2004;1:345e350. cancer patient. Int J Radiat Oncol Biol Phys 2005;63:25e36. 21 Wardman P. Electron transfer and oxidative stress as key 43 Harrison LB, Chadha M, Hill RJ, Hu K, Shasha D. Impact of factors in the design of drugs selectively active in hypoxia. tumor hypoxia and anemia on radiation therapy outcomes. Curr Med Chem 2001;8:739e761. Oncologist 2007;7:492e508. 22 Thomson P, Naylor MA, Everett SA, et al. Synthesis and 44 Hirst DG. What is the importance of anaemia in radiotherapy? biological properties of bioreductively targeted nitrothienyl The value of animal studies. Radiother Oncol 1991;20(Suppl. 1): prodrugs of combretastatin A-4. Mol Cancer Ther 2006;5: 29e33. 2886e2894. 45 Thomas GM. Raising hemoglobin: an opportunity for increasing 23 Cohen-Jonathon E, Bernhard EJ, McKenna WG. How does survival? Oncology 2002;63(Suppl. 2):19e28. radiation kill cells? Curr Opin Chem Biol 1999;3:77e83. 46 Rowinsky EK. Novel radiation sensitizers targeting tissue 24 Brown JM, Wilson WR. Exploiting tumour hypoxia in cancer hypoxia. Oncology 1999;13(Suppl. 5):61e70. treatment. Nat Rev Cancer 2004;4:437e447. 47 Kaanders JHAM, Bussink J, van der Kogel AJ. Clinical studies of 25 Janssen HL, Haustermans KM, Balm AJ, Begg AC. Hypoxia in hypoxia modification in radiotherapy. Semin Radiat Oncol head and neck cancer: how much, how important? Head Neck 2004;14:233e240. 2005;27:622e638. 48 Aapro M, Vaupel P. Erythropoietin: effects on life expectancy 26 Isa AY, Ward TH, West CML, Slevin NJ, Homer JJ. Hypoxia in in patients with cancer-related anaemia. Curr Med Res Opin head and neck cancer. Br J Radiol 2006;79:791e798. 2006;22(Suppl. 4):5e13. 27 Dendy PP, Wardman P. Hypoxia in biology and medicine d the 49 Calais G, Hirst DG. In situ tumour radiosensitization induced by legacy of L H Gray. Br J Radiol 2006;79:545e549. clofibrate administration: single dose and fractionated stud- 28 Whillans DW, Rauth AM. An experimental and analytical study ies. Radiother Oncol 1991;22:99e103. of oxygen depletion in stirred cell suspensions. Radiat Res 50 Abraham DJ, Wireko FC, Randad RS, et al. Allosteric modifiers 1980;84:97e114. of hemoglobin: 2-[4-[[(3,5-disubstituted anilino)carbonyl] 29 Koch CJ. A thin-film culturing technique allowing rapid methyl]phenoxy]-2-methylpropionic acid derivatives that lower gaseliquid equilibration (6 sec) with no toxicity to mammalian the oxygen affinity of hemoglobin in red cell suspensions, in cells. Radiat Res 1984;97:434e442. whole blood, and in vivo in rats. Biochemistry 1992;31: 30 Joiner MC, Marples B, Lambin P, Short SC, Turesson I. Low-dose 9141e9149. hypersensitivity: current status and possible mechanisms. Int J 51 Khandelwal SR, Kavanagh BD, Lin P-S, et al. RSR13, an Radiat Oncol Biol Phys 2001;49:379e389. allosteric effector of haemoglobin, and carbogen radiosensi- 31 Krause M, Wohlfarth J, Georgi B, et al. Low-dose hyper- tize FSAII and SCCVII tumours in C3H mice. Br J Cancer 1999; radiosensitivity of human glioblastoma cell lines in vitro does 79:814e820. not translate into improved outcome of ultrafractionated 52 Wahr JA, Gerber M, Venitz J, Baliga N. Allosteric modification radiotherapy in vivo. Int J Radiat Biol 2005;81:751e758. of oxygen delivery by hemoglobin. Anesth Analg 2001;92: 32 Palcic B, Brosing JW, Skarsgard LD. Survival measurements at low 615e620. doses: oxygen enhancement ratio. Br J Cancer 1982;46:980e984. 53 Mehta MP, Suh JH. Novel radiosensitizers for tumors of the 33 Watts ME, Hodgkiss RJ, Jones NR, Fowler JF. Radiosensitization central nervous system. Curr Opin Investig Drugs 2004;5: of Chinese hamster cells by oxygen and misonidazole at low 1284e1291. X-ray doses. Int J Radiat Biol 1986;50:1009e1021. 54 Rosenberg A, Knox S. Radiation sensitization with redox 34 Marples B, Joiner MC, Skov KA. The effect of oxygen on modulators: a promising approach. Int J Radiat Oncol Biol low-dose hypersensitivity and increased radioresistance in Phys 2006;64:343e354. Chinese hamster V79-379A cells. Radiat Res 1994;138: 55 Stea B, Suh JH, Boyd AP, Cagnoni PJ, Shaw E, Group RS. Whole- S17eS20. brain radiotherapy with or without efaproxiral for the 35 Wardman P, Rothkamm K, Folkes LK, Woodcock M, treatment of brain metastases: determinants of response Johnston PJ. Radiosensitization by nitric oxide at low radiation and its prognostic value for subsequent survival. Int J Radiat doses. Radiat Res 2007;167:475e484. Oncol Biol Phys 2006;64:1023e1030. 412 CLINICAL ONCOLOGY

56 Collingridge DR, Rockwell S. Pentoxifylline improves the 74 Adams GE. Hypoxia-mediated drugs for radiation and chemo- oxygenation and radiation response of BA1112 rat rhabdomyo- therapy. Cancer 1981;48:696e707. sarcomas and EMT6 mouse mammary carcinomas. Int J Cancer 75 Stewart FA, Denekamp J, Randhawa VS. Skin sensitization by 2000;90:256e264. misonidazole: a demonstration of uniform mild hypoxia. Br J 57 Nieder C, Zimmermann FB, Adam M, Molls M. The role of Cancer 1982;45:869e877. pentoxifylline as a modifier of radiation therapy. Cancer Treat 76 Hendry JH. Sensitization of hypoxic normal tissue. Br J Cancer Rev 2005;31:448e455. 1978;37(Suppl. III):232e234. 58 Kjellen E, Joiner MC, Collier JM, Johns H, Rojas A. A 77 Fowler JF. Chemical modifiers of radiosensitivity d theory and therapeutic benefit from combining normobaric carbogen or reality: a review. Int J Radiat Oncol Biol Phys 1985;11: oxygen with nicotinamide in fractionated X-ray treatments. 665e674. Radiother Oncol 1991;22:81e91. 78 Wouters BG, Brown JM. Cells at intermediate oxygen levels can 59 Horsman MR. Nicotinamide and the hypoxia problem. Radio- be more important than the ‘‘hypoxic fraction’’ in determining ther Oncol 1991;22:79e80. tumor response to fractionated radiotherapy. Radiat Res 1997; 60 Rojas A, Denekamp J, Johns H, et al. Nicotinamide as a repair 147:541e550. inhibitor in vivo: studies using single and fractionated X-ray 79 Overgaard J. Clinical evaluation of nitroimidazoles as modi- doses in mouse skin and kidneys. Radiat Res 1996;145:419e431. fiers of hypoxia in solid tumors. Oncol Res 1994;6:509e518. 61 Rojas A, Vojnovic B, Johns H, et al. Radiosensitization in 80 Adams GE, Clarke ED, Flockhart IR, et al. Structureeactivity normal tissues with oxygen, carbogen or nicotinamide: relationships in the development of hypoxic cell radiosensi- therapeutic gain comparisons for fractionated X-ray schedules. tizers: I. Sensitization efficiency. Int J Radiat Biol 1979;35: Radiother Oncol 1996;39:53e64. 133e150. 62 Hoskin PJ, Saunders MI. ARCON. Clin Oncol (R Coll Radiol) 81 Adams GE, Clarke ED, Gray P, et al. Structureeactivity 1994;6:281e282. relationships in the development of hypoxic cell radiosensi- 63 Saunders MI, Hoskin PJ, Pigott K, et al. Accelerated radio- tizers: II. Cytotoxicity and therapeutic ratio. Int J Radiat Biol therapy, carbogen and nicotinamide (ARCON) in locally 1979;35:151e160. advanced head and neck cancer: a feasibility study. Radiother 82 Coleman CN, Wasserman TH, Urtasun RC, et al. Final report of Oncol 1997;45:159e166. the phase I trial of the hypoxic cell radiosensitizer SR 2508 64 Hoskin PJ, Saunders MI, Phillips H, et al. Carbogen and (etanidazole) Radiation Therapy Oncology Group 83-03. Int J nicotinamide in the treatment of bladder cancer with radical Radiat Oncol Biol Phys 1990;18:389e393. radiotherapy. Br J Cancer 1997;76:260e263. 83 Smithen CE, Clarke ED, Dale JA, et al. Novel (nitro-1- 65 Kaanders JHAM, Stratford MRL, Liefers J, et al. Administration imidazolyl) alkanolamines as potential radiosensitizers with of nicotinamide during a five to seven week course of improved therapeutic properties. In: Brady LW, editor. radiotherapy: pharmcokinetics, tolerance, and compliance. Radiation sensitizers: their use in the clinical management Radiother Oncol 1997;43:67e73. of cancer. New York: Masson, 1980:22e32. 66 Hoskin PJ, Rojas AM, Phillips H, Saunders MI. Acute and late 84 Dennis MF, Stratford MRL, Wardman P, Watts ME. Cellular morbidity in the treatment of advanced bladder carcinoma uptake of misonidazole and analogues with acidic or basic with accelerated radiotherapy, carbogen, and nicotinamide. functions. Int J Radiat Biol 1985;47:629e643. Cancer 2005;103:2287e2297. 85 Watts ME, Dennis MF, Roberts IJ. Radiosensitization by 67 Kaanders JHAM, Bussink J, van der Kogel AJ. ARCON: a novel misonidazole, pimonidazole and azomycin and intracellular biology-based approach in radiotherapy. Lancet Oncol 2002;3: uptake in human tumour cell lines. Int J Radiat Biol 1990;57: 728e737. 361e372. 68 Bernier J, Denekamp J, Rojas A, et al. ARCON: accelerated 86 Dische S, Bennett MH, Orchard R, Stratford MRL, Wardman P. radiotherapy with carbogen and nicotinamide in head and The uptake of the radiosensitizing compound Ro 03-8799 neck squamous cell carcinomas. The experience of the (pimonidazole) in human tumors. Int J Radiat Oncol Biol Phys Cooperative Group of Radiotherapy of the European Organi- 1989;16:1089e1092. zation for Research and Treatment of Cancer (EORTC). 87 Raleigh JA, Dewhirst MW, Thrall DE. Measuring tumor hypoxia. Radiother Oncol 2000;55:111e119. Semin Radiat Oncol 1996;6:37e45. 69 Hoogsteen IJ, Pop LAM, Marres HAM, et al. Oxygen-modifying 88 Overgaard J, Eriksen JG, Nordsmark M, Alsner J, Horsman MR, treatment with ARCON reduces the prognostic significance of Group DHaNCS. Plasma osteopontin, hypoxia, and response to hemoglobin in squamous cell carcinoma of the head and neck. the hypoxia sensitiser nimorazole in radiotherapy of head Int J Radiat Oncol Biol Phys 2006;64:83e89. and neck cancer: results from the DAHANCA 5 randomised 70 Adams GE, Cooke MS. Electron-affinic sensitization. I. A double-blind placebo-controlled trial. Lancet Oncol 2005;6: structural basis for chemical radiosensitizers in bacteria. Int 757e764. J Radiat Biol 1969;15:457e471. 89 Nagasawa H, Uto Y, Kirk KL, Hori H. Design of hypoxia- 71 Adams GE, Asquith JC, Dewey DL, Foster JL, Michael BD, targeting drugs as new cancer chemotherapeutics. Biol Pharm Willson RL. Electron affinic sensitization. Part II: para- Bull 2006;29:2335e2342. nitroacetophenone: a radiosensitizer for anoxic bacterial and 90 Sugie C, Shibamoto Y, Ito M, et al. Reevaluation of the mammalian cells. Int J Radiat Biol 1971;19:575e585. radiosensitizing effects of sanazole and nimorazole in vitro 72 Chapman JD, Webb RG, Borsa J. Radiosensitization of and in vivo. J Radiat Res (Tokyo) 2005;46:453. mammalian cells by p-nitroacetophenone. I. Characterization 91 Dobrowsky W, Huigol NG, Jayatilake RS, et al. AK-2123 in asynchronous and synchronous populations. Int J Radiat Biol (Sanazol) as a radiation sensitizer in the treatment of stage 1971;19:561e573. III cervical cancer: results of an IAEA multicentre randomised 73 Asquith JC, Foster JL, Willson RL, Ings R, McFadzean JA. trial. Radiother Oncol 2007;82:24e29. Metronidazole (‘Flagyl’). A radiosensitizer of hypoxic cells. 92 Das T, Chakraborty S, Banerjee S, et al. Preparation and Br J Radiol 1974;47:474e481. preliminary biological evaluation of a 177Lu labeled sanazole CHEMICAL RADIOSENSITIZERS IN RADIOTHERAPY 413

derivative for possible use in targeting tumor hypoxia. Bioorg 111 Hicks KO, Pruijn FB, Secomb TW, et al. Use of three- Med Chem 2004;12:6077e6084. dimensional tissue cultures to model extravascular transport 93 Jonathan RA, Wijffels KIEM, Peeters W, et al. The prognostic and predict in vivo activity of hypoxia-targeted anticancer value of endogenous hypoxia-related markers for head and drugs. J Natl Cancer Inst 2006;98:1118e1128. neck squamous cell carcinomas treated with ARCON. Radio- 112 Minchinton AI, Tannock IF. Drug penetration in solid tumours. ther Oncol 2006;79:288e297. Nat Rev Cancer 2006;6:583e592. 94 Koch CJ, Evans SM, Lord EM. Oxygen dependence of cellular 113 Brown JM. Therapeutic targets in radiotherapy. Int J Radiat uptake of EF5 [2-(2-nitro-1H-imidazol-1-yl)-N-(2,2,3,3,3- Oncol Biol Phys 2001;49:319e326. pentafluoropropyl) acetamide]: analysis of drug adducts by 114 Peters KB, Brown JM. Tirapazamine: a hypoxia-activated fluorescent antibodies vs. bound radioactivity. Br J Cancer topoisomerase II poison. Cancer Res 2002;62:5248e5253. 1995;72:869e874. 115 Foster JL, Conroy PJ, Searle AJ, Willson RL. Metronidazole 95 Chapman JD, Schneider RF, Urbain JL, Hanks GE. Single-photon (Flagyl): characterization as a cytotoxic drug specific for emission computed tomography and positron-emission tomo- hypoxic tumour cells. Br J Cancer 1976;33:485e490. graphy assays for tissue oxygenation. Semin Radiat Oncol 116 Wardman P, Clarke ED. Oxygen inhibition of nitroreductase: 2001;11:47e57. electron transfer from nitro radical-anions to oxygen. Biochem 96 Sunamura M, Karasawa K, Okamoto A, et al. Phase III trial of Biophys Res Commun 1976;69:942e949. radiosensitizer PR-350 combined with intraoperative radio- 117 Laderoute K, Wardman P, Rauth AM. Molecular mechanisms for therapy for the treatment of locally advanced pancreatic the hypoxia-dependent activation of 3-amino-1,2,4-benzo- cancer. Pancreas 2004;28:330e334. triazine-1,4-dioxide (SR 4233). Biochem Pharmacol 1988;37: 97 von Sonntag C. The chemical basis of radiation biology. 1487e1495. London: Taylor & Francis, 1987. 118 Brown JM, Lemmon MJ. Potentiation of the hypoxic cytotoxin 98 von Sonntag C. Free-radical-induced DNA damage and its SR 4233 of cell killing produced by fractionated irradiation of repair. A chemical perspective. Berlin: Springer, 2006. mouse tumors. Cancer Res 1990;50:7745e7749. 99 Wardman P. Sensitization and protection of oxidative damage 119 Priyadarsini KI, Tracy M, Wardman P. The one-electron caused by high energy radiation. In: Scott G, editor. reduction potential of 3-amino-1,2,4-benzotriazine 1,4- Atmospheric oxidation and antioxidants, Vol. III. Amsterdam: dioxide (tirapazamine): a hypoxia-selective bioreductive drug. Elsevier, 1993:101e127. Free Radic Res 1996;25:393e399. 100 Bamatraf MMM, O’Neill P, Rao BSM. Redox dependence of the 120 Wardman P. Chemistry of nitroarene and aromatic N-oxide rate of interaction of hydroxyl radical adducts of DNA radicals. In: Alfassi ZB, editor. The chemistry of N-centered nucleobases with oxidants: consequences for DNA strand radicals. New York: Wiley, 1998:615e661. breakage. J Am Chem Soc 1998;120:11852e11857. 121 Zeman EM, Brown JM. Pre- and post-irradiation radiosensiti- 101 Skov KA. Workshop report. DNA targeted hypoxic cytotoxins zation by SR 4233. Int J Radiat Oncol Biol Phys 1989;16: and radiosensitizers. Int J Radiat Biol 1989;56:387e393. 967e971. 102 Cowan DSM, Matejovic JF, McClelland RA, Rauth AM. DNA- 122 Zeman EM, Lemmon MJ, Brown JM. Aerobic radiosensitization targeted 2-nitroimidazoles: in vitro and in vivo studies. Br J by SR 4233 in vitro and in vivo. Int J Radiat Oncol Biol Phys Cancer 1994;70:1067e1074. 1990;18:125e132. 103 Cowan DSM, Matejovic JF, Wardman P, McClelland RA, 123 Itani W, Geara F, Haykal J, Haddadin M, Gali-Muhtasib H. Rauth AM. Radiosensitizing and cytotoxic properties of DNA- Radiosensitization by 2-benzoyl-3-phenyl-6,7-dichloroquinoxa- targeted phenanthridine-linked nitroheterocycles of varying line 1,4-dioxide under oxia and hypoxia in human colon cancer electron affinities. Int J Radiat Biol 1994;66:729e738. cells. Radiat Oncol 2007;2:1e13. 104 Wilson WR, Siim BG, Denny WA, et al. 5-Nitro-4-(N,N- 124 Jenkins TC, Naylor MA, O’Neill P, et al. Synthesis and dimethylaminopropylamino)quinoline (5-nitraquine), a new evaluation of a-[[(2-haloethyl)amino]methyl]-2-nitro-1H- DNA-affinic hypoxic cell radiosensitizer and bioreductive imidazole-1-ethanols as prodrugs of a-[(1-aziridinyl)methyl]2- agent: comparison with nitracrine. Radiat Res 1992;131: nitro-1H-imidazole-1-ethanol (RSU-1069) and its analogues 257e265. which are radiosensitizers and bioreductively activated cyto- 105 Parrick J, Porssa M, Davies LK, et al. Targeting radiosensitizers toxins. J Med Chem 1990;33:2603e2610. to DNA by minor groove binding: nitroarenes based on 125 Suto MJ, Stier MA, Werbel LM, et al. A new class of analogues netropsin and distamycin. Bioorg Med Chem Lett 1993;3: of the bifunctional radiosensitizer a-(1-aziridinylmethyl)2- 1697e1702. nitro-1H-imidazole-1-ethanol (RSU 1069): the cycloalkylazir- 106 Holley J, Mather A, Cullis P, et al. Uptake and cytotoxicity of idines. J Med Chem 1991;34:2484e2488. novel nitroimidazoleepolyamine conjugates in Ehrlich ascites 126 Bremner JCM. Assessing the bioreductive effectiveness of the tumour cells. Biochem Pharmacol 1992;43:763e769. nitroimidazole RSU1069 and its prodrug RB6145: with partic- 107 Wardman P, Dennis MF, White J. A probe for intracellular ular reference to in vivo methods of evaluation. Cancer concentrations of drugs: delayed fluorescence from acridine Metastasis Rev 1993;12:177e193. orange. Int J Radiat Oncol Biol Phys 1989;16:935e938. 127 Wood PJ, Horsman MR, Khalil AA, et al. A comparison of the 108 Stratford MRL, Dennis MF, Watts ME, Watfa RR, Woodcock M. physiological effects of RSU1069 and RB6145 in the SCCVII Radiosensitizer-DNA interactions in relation to intracellular murine tumour. Acta Oncol 1996;35:989e994. uptake. Int J Radiat Oncol Biol Phys 1989;16:1007e1010. 128 Lunt SJ, Telfer BA, Fitzmaurice RJ, Stratford IJ, Williams KJ. 109 Wilson WR, Hicks KO. Measurement of extravascular drug Tirapazamine administered as a neoadjuvant to radiotherapy diffusion in muticellular layers. Br J Cancer 1999;79:1620e1628. reduces metastatic dissemination. Clin Cancer Res 2005;11: 110 Kyle AH, Minchinton AI. Measurement of delivery and 4212e4216. metabolism of tirapazamine to tumour tissue using the 129 Ross AB, Mallard WG, Helman WP, Buxton GV, Huie RE, Neta P. mutilayered cell culture model. Cancer Chemother Pharmacol NDRL-NIST Solution Kinetics Database: Version 3. Notre Dame, 1999;43:213e220. Indiana and Gaithersburg, Maryland: Notre Dame Radiation 414 CLINICAL ONCOLOGY

Laboratory and National Institute of Standards and Tech- 148 Sonveaux P, Feron O. Nitric oxide and tumor biology. In: nology, 1998. Lamas S, Cadenas E, editors. Nitric oxide, cell signaling, and 130 Shafirovich V, Lymar SV. Nitroxyl and its anion in aqueous gene expression. Boca Raton: CRC Press, 2006:395e420. solutions: spin states, protic equilibria, and reactivities toward 149 Ichinose F, Roberts JD Jr, Zapol WM. Inhaled nitric oxide. A oxygen and nitric oxide. Proc Natl Acad Sci USA 2002;99: selective pulmonary vasodilator. Current uses and therapeutic 7340e7345. potential. Circulation 2004;109:3106e3111. 131 Whitaker SJ, McMillan TJ. Oxygen effect for DNA double-strand 150 Hataishi R, Zapol WM, Bloch KD, Ichinose F. Inhaled nitric oxide break induction determined by pulsed-field gel electrophore- does not reduce systemic vascular resistance in mice. Am J sis. Int J Radiat Biol 1992;61:29e41. Physiol Heart Circ Physiol 2006;209:H1826eH1829. 132 Olive PL, Ba´nath JP, MacPhail HS. Lack of correlation between 151 Ng Q-S, Goh V, Milner J, et al. Effect of nitric-oxide synthesis radiosensitivity and DNA double-strand breaks: induction or on tumour blood volume and vascular activity: a phase I study. rejoining in six human tumor cell lines. Cancer Res 1994;54: Lancet Oncol 2007;8:111e118. 3939e3946. 152 Siemann DW, Chaplin DJ, Horsman MR. Vascular-targeting 133 Howard-Flanders P. Effect of nitric oxide on the radiosensitiv- therapies for treatment of malignant disease. Cancer 2004; ity of bacteria. Nature 1957;180:1191e1192. 100:2491e2499. 134 Gray LH, Green FO, Hawes CA. Effect of nitric oxide on the 153 Tozer GM, Kanthou C, Baguley BC. Disrupting tumour blood radiosensitivity of tumour cells. Nature 1958;182:952e953. vessels. Nat Rev Cancer 2005;5:423e435. 135 Mitchell JB, Wink DA, DeGraff W, Gamson J, Keefer LK, 154 Denekamp J. Limited role of vasculature-mediated injury in Krishna MC. Hypoxic mammalian cell radiosensitization by tumor response to radiotherapy. J Natl Cancer Inst 1994;85: nitric oxide. Cancer Res 1993;53:5845e5848. 935e937. 136 Mitchell JB, Cook JA, Krishna MC, et al. Radiation sensitization 155 Tozer GM, Everett SA. Nitric oxide in tumour biology and by nitric oxide releasing agents. Br J Cancer 1996;74(Suppl. cancer therapy. Part 2: therapeutic implications. Clin Oncol XXVII):S181eS184. (R Coll Radiol) 1997;9:357e364. 137 Mitchell JB, DeGraff W, Kim S, et al. Redox generation of nitric 156 Tozer GM, Everett SA. Nitric oxide in tumour biology and oxide to radiosensitize hypoxic cells. Int J Radiat Oncol Biol cancer therapy. Part 1: physiological aspects. Clin Oncol (R Phys 1998;42:795e798. Coll Radiol) 1997;9:282e293. 138 Ignarro LJ, editor. Nitric oxide. Biology and pathobiology. San 157 Jordan BF, Gre´goire V, Demeure RJ, et al. Insulin increases the Diego: Academic Press, 2000. sensitivity of tumors to irradiation: involvement of an increase 139 Worthington J, McCarthy HO, Barrett E, Adams C, Robson T, in tumor oxygenation mediated by a nitric oxide-dependent Hirst DG. Use of the radiation-inducible WAF1 promoter to decrease of the tumor cells oxygen consumption. Cancer Res drive iNOS gene therapy as a novel anti-cancer treatment. 2002;62:3555e3561. J Gene Med 2004;6:673e680. 158 De Ridder M, Verovski VN, Chiavaroli C, et al. The radio- 140 Wang Z, Cook T, Alber S, et al. Adenoviral gene transfer of the sensitizing effect of immunoadjuvant OM-174 requires co- human inducible nitric oxide synthase gene enhances the operation between immune and tumor cells through radiation response of human colorectal cancer associated with interferon-gamma and inducible nitric oxide synthase. Int J alterations in tumor vascularity. Cancer Res 2004;64:1386e1395. Radiat Oncol Biol Phys 2006;66:1473e1480. 141 Cook T, Wang Z, Alber S, et al. Nitric oxide and ionizing 159 Behar D, Neta P. Intramolecular electron transfer and radiation synergistically promote apoptosis and growth in- dehalogenation of anion radicals. 2. Halonitroaromatic com- hibition of cancer by activating p53. Cancer Res 2004;64: pounds. J Phys Chem 1981;85:690e693. 8015e8021. 160 Wilson WR, Ferry DM, Tercel M, Anderson RF, Denny WA. 142 Worthington J, Robson T, Scott S, Hirst D. Evaluation of Reduction of nitroarylmethyl quaternary ammonium prodrugs a synthetic CArG promoter for nitric oxide synthase gene of mechloramine by radiation. Radiat Res 1998;149:237e245. therapy of cancer. Gene Ther 2005;12:1417e1423. 161 Kriste AG, Tercel M, Anderson RF, Ferry DM, Wilson WR. 143 McCarthy HO, Worthington J, Barrett E, et al. p21(WAF1)- Pathways of reductive fragmentation of heterocyclic nitro- mediated transcriptional targeting of inducible nitric oxide arylmethyl quaternary ammonium prodrugs of mechloreth- synthase gene therapy sensitizes tumours to fractionated amine. Radiat Res 2002;158:753e762. radiotherapy. Gene Ther 2007;14:246e255. 162 Ahn G-O, Ware DC, Denny WA, Wilson WR. Optimization of the 144 Jordan BF, Beghein N, Aubry M, Gre´goire V, Gallez B. auxiliary ligand shell of cobalt(III)(8-hydroxyquinoline) com- Potentiation of radiation-induced regrowth delay by isosorbide plexes as model hypoxia-selective radiation-activated pro- dinitrate in FSAII murine tumors. Int J Cancer 2003;103: drugs. Radiat Res 2004;162:315e325. 138e141. 163 Ahn G-O, Botting KJ, Patterson AV, Ware DC, Tercel M, 145 Jordan BF, Sonveaux P, Feron O, Gre´goire V, Beghein N, Wilson WR. Radiolytic and cellular reduction of a novel Gallez B. Nitric oxide-mediated increase in tumor blood flow hypoxia-activated cobalt(III) prodrug of a chloromethylbenz- and oxygenation of tumors implanted in muscles stimulated by indoline DNA minor groove alkylator. Biochem Pharmacol 2006; electric pulses. Int J Radiat Oncol Biol Phys 2003;55: 71:1683e1694. 1066e1073. 164 Tanabe K, Makimura Y, Tachi Y, Imagawa-Sato A, Nishimoto S. 146 Kashiwagi S, Izumi Y, Gohongi T, et al. NO mediates mural cell Hypoxia-selective activation of 5-fluorodeoxyuridine prodrug recruitment and vessel morphogenesis in murine melanomas possessing indolequinone structure: radiolytic reduction and and tissue-engineered blood vessels. J Clin Invest 2005;115: cytotoxicity characteristics. Bioorg Med Chem Lett 2005;15: 1816e1827. 2321e2324. 147 Shan SQ, Rosner GL, Braun RD, Hahn J, Pearce C, Dewhirst MW. 165 Shibamoto Y, Tachi T, Tanabe K, Hatta H, Nishimoto S-I. Effects of diethylamine/nitric oxide on blood perfusion and In vitro and in vivo evaluation of novel antitumor prodrugs of oxygenation in the R3230Ac mammary carcinoma. Br J Cancer 5-fluoro-2-deoxyuridine activated by hypoxic irradiation. Int 1997;76:429e437. J Radiat Oncol Biol Phys 2004;58:397e402. CHEMICAL RADIOSENSITIZERS IN RADIOTHERAPY 415

166 Cramp WA. Radiosensitization by copper ions, and consequent mammary tumors. Int J Radiat Oncol Biol Phys 2004;58: reversal of the oxygen effect. Nature 1965;206:636e637. 1570e1576. 167 Hesslewood IP, Cramp WA, McBrien DCH, Williamson P, 186 Magda D, Miller RA. Motexafin gadolinium: a novel redox active Lott KAK. Copper as a hypoxic cell sensitizer of mammalian drug for cancer therapy. Semin Cancer Biol 2006;16:466e476. cells. Br J Cancer 1978;37(Suppl. III):95e97. 187 Khuntia D, Mehta M. Motexafin gadolinium: a clinical review of 168 Selvaraj S, Chabita (Saha) K, Chakraborty A, Bhattacharyya SN, a novel radioenhancer for brain tumors. Expert Rev Anticancer Saha A. Influence of copper (II) ions and its derivatives on Ther 2004;4:981e989. radiosensitivity of Escherichia coli. Radiat Phys Chem (in press). 188 Richards GM, Mehta MP. Motexafin gadolinium in the treatment of 169 Buettner GR, Jurkiewicz BA. Catalytic metals, ascorbate and brain metastases. Expert Opin Pharmacother 2007;8:351e359. free radicals: combinations to avoid. Radiat Res 1996;145: 189 Gomer CJ, Rucker N, Ferrario A, Wong S. Properties and 532e541. applications of photodynamic therapy. Radiat Res 1989;120: 170 Burkitt MJ. A critical overview of the chemistry of copper- 1e18. dependent low density lipoprotein oxidation: roles of lipid 190 Schaffer M, Schaffer PM, Hofstetter A, Du¨hmke E, Jori G. On hydroperoxides, alpha-tocopherol, thiols, and ceruloplasmin. the double role of Photofrin as a photo- and a radio-sensitising Arch Biochem Biophys 2001;394:117e135. agent: a possible new combination therapy for tumours. 171 Antholine WE, Byrnes R, Petering DH. Copper complexes as Photochem Photobiol 2002;1:438e439. drugs. In: Berthon G, editor. Handbook of metaleligand 191 Schaffer M, Ertl-Wagner B, Schaffer PM, et al. The application interactions in biological fluids, Vol. 2. New York: Marcel of Photofrin II as a sensitizing agent for ionizing radiation d Dekker, 1995:1024e1038. a new approach in tumor therapy? Curr Med Chem 2005;12: 172 Wardman P. The kinetics of the reaction of ‘anomalous’ 4- 1209e1215. nitroimidazole radiosensitizers with thiols. Int J Radiat Biol 192 Schaffer M, Ertl-Wagner B, Schaffer PM, et al. Feasibility of 1982;41:231e235. photofrin II as a radiosensitizing agent in solid tumors d pre- 173 Wardman P, Clarke ED. Redox properties and rate constants in liminary results. Onkologie 2006;29:514e519. free-radical mediated damage. Br J Cancer 1987;55(Suppl. 193 Kinsella TJ, Dobson PP, Mitchell JB, Fornace AJ Jr. Enhance- VIII):172e177. ment of X-ray induced DNA damage by prep-treatment with 174 Young SW, Qing F, Harriman A, et al. Gadolinium(III) halogenated pyrimidine analogs. Int J Radiat Oncol Biol Phys texaphyrin: a tumor selective radiation sensitizer that is 1987;13:1e7. detectable by MRI. Proc Natl Acad Sci USA 1996;93:6610e 194 Seo Y, Yan T, Schupp JE, Colussi V, Taylor KL, Kinsella TJ. 6615. Differential radiosensitization in DNA mismatch repair- 175 Bernhard EJ, Mitchell JB, Deen D, Cardell M, Rosenthal DI, proficient and -deficient human colon cancer xenografts with Brown JM. Re-evaluating gadolinium(III) texaphyrin as a radio- 5-iodo-2-pyrimidinone-2-deoxyribose. Clin Cancer Res 2004; sensitizing agent. Cancer Res 2000;60:86e91. 10:7520e7528. 176 Sessler JL, Tvermoes NA, Guldi DM, Mody TD, Allen WE. One- 195 Berry SE, Kinsella TJ. Targeting DNA mismatch repair for electron reduction and oxidation studies of the radiation radiosensitization. Semin Radiat Oncol 2001;11:300e315. sensitizer gadolinium(III) texaphyrin (PCI-0120) and other 196 McGinn CJ, Shewach DS, Lawrence TS. Radiosensitizing water soluble metallotexaphyrins. J Phys Chem A 1999;103: nucleosides. J Natl Cancer Inst 1996;88:1193e1203. 787e794. 197 McGinn CJ, Lawrence TS. Recent advances in the use of radio- 177 Magda D, Lepp C, Gerasimchuk N, et al. Redox cycling by sensitizing nucleosides. Semin Radiat Oncol 2001;11:270e280. motexafin gadolinium enhances cellular response to ionizing 198 Glynne-Jones R, Debus J. Improving chemoradiotherapy in radiation by forming reactive oxygen species. Int J Radiat rectal cancer. Oncologist 2001;6(Suppl. 4):29e34. Oncol Biol Phys 2001;51:1025e1036. 199 Zhu AX, Willett CG. Chemotherapeutic and biologic agents as 178 Wood PM. The redox potential of the system oxygen- radiosensitizers in rectal cancer. Semin Radiat Oncol 2003;13: superoxide. FEBS Lett 1974;44:21e24. 454e468. 179 Wardman P. The reduction potential of benzyl viologen: an 200 Vallerga AK, Zarling DA, Kinsella TJ. New radiosensitizing important reference compound for oxidant/radical redox regimens, drugs, prodrugs, and candidates. Clin Adv Hematol couples. Free Radic Res Commun 1991;14:57e67. Oncol 2004;2:793e805. 180 Renschler MF. The emerging role of reactive oxygen species in 201 Pauwels B, Korst AEC, Lardon F, Vermorken JB. Combined cancer therapy. Eur J Cancer 2004;40:1934e1940. modality therapy of gemcitabine and radiation. Oncologist 181 Mikkelsen RB, Wardman P. Biological chemistry of reactive 2005;10:34e51. oxygen and nitrogen and ionizing radiation-induced signal trans- 202 Douple EB, Richmond RC, O’Hara JA, Coughlin CT. Carboplatin duction mechanisms. Oncogene 2003;22:5734e5754. as a potentiator of radiation therapy. Cancer Treat Rev 1985; 182 Rockwell S, Donnelly ET, Liu Y, Tang LQ. Preliminary studies of 12(Suppl. A):111e124. the effects of gadolinium texaphyrin on the growth and 203 Yang L, Douple EB, O’Hara JA, Crabtree RA, Eastman A. radiosensitivity of EMT6 cells in vitro. Int J Radiat Oncol Biol Enhanced radiation-induced cell killing by carboplatin in cells Phys 2002;54:536e541. of repair-proficient and repair-deficient cell lines. Radiat Res 183 Donnelly ET, Liu Y, Paul TK, Rockwell S. Effects of motexafin 1995;144:230e236. gadolinium on DNA damage and X-ray-induced DNA damage 204 Yang LX, Douple EB, O’Hara JA, Wang HJ. Production of DNA repair, as assessed by the Comet assay. Int J Radiat Oncol Biol double-strand breaks by interactions between carboplatin and Phys 2004;62:1176e1186. radiation: a potential mechanism for radiopotentiation. Radiat 184 Xu S, Zakian K, Thaler H, et al. Effects of motexafin gadolinium Res 1995;143:309e315. on tumor metabolism and radiation sensitivity. Int J Radiat 205 Skov KA, Farrell NP, Adomat H. Platinum complexes with one

Oncol Biol Phys 2001;49:1381. radiosensitizing ligand [PtCl2(NH3)(Sensitizer)]: radiosensitiza- 185 Donnelly ET, Liu Y, Fatunmbi YO, Lee I, Magda D, Rockwell S. tion and toxicity studies in vitro. Radiat Res 1987;112: Effects of texaphyrins on the oxygenation of EMT6 mouse 273e282. 416 CLINICAL ONCOLOGY

206 Skov KA. Modification of radiation response by metal com- 227 Williams KJ, Telfer BA, Brave S, et al. ZD6474, a potent plexes: a review with emphasis on nonplatinum studies. Radiat inhibitor of vascular endothelial growth factor signaling, Res 1987;112:217e242. combined with radiotherapy: schedule-dependent enhance- 207 Chan PKL, Skov KA, James BR. Further studies on toxic ment of antitumor activity. Clin Cancer Res 2004;10: and radiosensitizing properties of ruthenium complexes of 8587e8593. 4-nitroimidazoles. Int J Radiat Biol 1987;52:49e55. 228 Bonner JA, Harari PM, Giralt J, et al. Radiotherapy plus 208 Eberhardt W, Po¨ttgen C, Stuschke M. Chemoradiation para- cetuximab for squamous-cell carcinoma of the head and neck. digm for the treatment of lung cancer. Nat Clin Pract Oncol N Engl J Med 2006;354:567e578. 2006;3:188e199. 229 Posner MR, Wirth LJ. Cetuximab and radiotherapy for head and 209 Hao D, Ritter MA, Oliver T, Browman GP. Platinum-based neck cancer. N Engl J Med 2006;354:634e636. concurrent chemoradiotherapy for tumors of the head and 230 Harari PM, Huang S. Radiation combined with EGFR signal neck and the esophagus. Semin Radiat Oncol 2006;18: inhibitors: head and neck cancer focus. Semin Radiat Oncol 10e19. 2006;16:38e44. 210 Miyamoto S, Huang TT, Wu¨rzberger-Davis S, et al. Cellular and 231 Liao Z, Komaki R, Milas L, et al. A phase I clinical trial of molecular responses to topoisomerase I poisons. Exploiting thoracic radiotherapy and concurrent celecoxib for patients synergy for improved radiotherapy. Ann N Y Acad Sci 2000;922: with unfavorable performance status inoperable/unresectable 274e292. non-small cell lung cancer. Clin Cancer Res 2005;11: 211 Chen AY, Chou R, Shih S-J, Laub D, Gandara D. Enhancement of 3342e3348. radiotherapy with DNA topoisomerase I-targeted drugs. Crit 232 Raju U, Ariga H, Dittmann K, Nakata E, Ang KK, Milas L. Rev Oncol Hematol 2004;50:111e119. Inhibition of DNA repair as a mechanism of enhanced radio- 212 Shih SJ, Erbele T, Chen AY. Ku86 modulates DNA topoisomerase response of head and neck carcinoma cells by a selective I-mediated radiosensitization, but not cytotoxicity, in mam- cyclooxygenase-2 inhibitor, celecoxib. Int J Radiat Oncol Biol malian cells. Cancer Res 2005;65:9194e9199. Phys 2005;63:520e528. 213 Wong ET, Berkenblit A. The role of topotecan in the treatment 233 Nieder C, Wiedenmann N, Andratschke N, Molls M. Current of brain metastases. Oncologist 2004;9:68e79. status of angiogenesis inhibitors combined with radiation 214 Patterson LH, McKeown SR, Ruparelia K, et al. Enhancement of therapy. Cancer Treat Rev 2006;32:348e364. chemotherapy and radiotherapy of murine tumours by AQ4N, 234 Miura M, Sakimoto I, Ohta K, Sugawara F, Sakaguchi K. a bioreductively activated anti-tumour agent. Br J Cancer Sulfoglycolipids as candidate antiangiogenic radiosensitizers. 2000;82:1984e1990. Anti-Cancer Drugs 2007;18:1e5. 215 McErlane V, Yakkundi A, McCarthy HO, et al. A cytochrome 235 Jendrossek V, Handrick R. Membrane targeted anticancer P450 2B6 meditated gene therapy strategy to enhance the drugs: potent inducers of apoptosis and putative radiosensi- effects of radiation or cyclophosphamide when combined tizers. Curr Med Chem Anti-Cancer Agents 2003;3:343e353. with the bioreductive drug AQ4N. J Gene Med 2005;7: 236 Bridges BA. Sensitization of Escherichia coli to gamma- 851e859. radiation by N-ethylmaleimide. Nature 1960;4748:415. 216 Zhang M, Chakravarti A. Novel radiation-enhancing agents in 237 Bridges BA. Sensitization of organisms to radiation by malignant gliomas. Semin Radiat Oncol 2006;16:29e37. sulfhydryl-binding agents. Adv Radiat Biol 1969;3:123e176. 217 Vira´g L, Szabo´ C. The therapeutic potential of poly(ADP-ribose) 238 Watts ME, Whillans DW, Adams GE. Studies of the mechanisms polymerase inhibitors. Pharmacol Rev 2002;54:375e429. of radiosensitization of bacterial and mammalian cells by 218 Calabrese CR, Almassy R, Barton S, et al. Anticancer chemo- diamide. Int J Radiat Biol 1975;27:259e270. sensitization and radiosensitization by the novel poly(ADP- 239 Harris JW. Mammalian cell studies with diamide. Pharmacol ribose) polymerase-1 inhibitor AG14361. J Natl Cancer Inst Ther 1979;7:375e391. 2004;96:56e67. 240 Wardman P. Thiol reactivity towards drugs and radicals: some 219 Sarkaria JN, Eshleman JS. ATM as a target for novel radio- implications in the radiotherapy and chemotherapy of cancer. sensitizers. Semin Radiat Oncol 2001;11:316e327. In: Chatgilialoglu C, Asmus K-D, editors. Sulfur-centered 220 Wong JS, Ara G, Keyes SR, Herbst R, Coleman CN, Teicher BA. reactive intermediates in chemistry and biology. New York: Lisofylline as a modifier of radiation therapy. Oncol Res 1996; Plenum Press, 1990:415e427. 8:513e518. 241 Griffith OW, Meister A. Potent and specific inhibition of 221 Choudhury A, Cuddihy A, Bristow RG. Radiation and new glutathione synthesis by buthionine sulfoximine (S-n-butyl molecular agents. Part I: Targeting ATM-ATR checkpoints, DNA homocysteine sulfoximine). J Biol Chem 1979;254: repair, and the proteasome. Semin Radiat Oncol 2006;16: 7558e7560. 51e58. 242 Hodgkiss RJ, Middleton RW. Enhancement of misonidazole 222 Chinnaiyan P, Allen GW, Harari PM. Radiation and new radiosensitization by an inhibitor of glutathione biosynthesis. molecular agents. Part II: Targeting HDAC, HSP90, IGF-1R, Int J Radiat Biol 1983;43:179e183. PI3K, and Ras. Semin Radiat Oncol 2006;16:59e64. 243 Minchinton AI, Rojas A, Smith KA, et al. Glutathione depletion 223 Downward J. Targeting RAS signalling pathways in cancer in tissues after administration of buthionine sulphoximine. Int therapy. Nat Rev Cancer 2003;3:11e22. J Radiat Oncol Biol Phys 1984;10:1261e1264. 224 Gupta AK, Bakanauskas VJ, Cerniglia GJ, et al. The Ras 244 Rojas A, Smith KA, Soranson JA, Minchinton AI, Middleton RW, radiation resistance pathway. Cancer Res 2001;61:4278e4282. Denekamp J. Enhancement of misonidazole radiosensitization 225 Cengel KA, McKenna WG. Molecular targets for altering by buthionine sulphoximine. Radiother Oncol 1984;2: radiosensitivity: lessons from Ras as a pre-clinical and clinical 325e332. model. Crit Rev Oncol Hematol 2005;55:103e116. 245 Lespinasse F, Biscay P, Malaise EP, Guichard M. SR-2508 plus 226 Sartor CI. Mechanisms of disease: radiosensitization by buthionine sulfoximine or SR-2508 alone: effect on the epidermal growth factor receptor inhibitors. Nat Clin Pract radiation response and the glutathione content of a human Oncol 2004;1:80e87. tumor xenograft. Radiat Res 1987;110:149e154. CHEMICAL RADIOSENSITIZERS IN RADIOTHERAPY 417

246 Ono K, Komuro C, Nishidai T, et al. Radiosensitizing effect of 253 Koch CJ, Evans SM. Cysteine concentrations in rodent tumors: misonidazole in combination with an inhibitor of glutathione unexpectedly high values may cause therapy resistance. Br J synthesis in murine tumors. Int J Radiat Oncol Biol Phys 1986; Cancer 1996;67:661e667. 12:1661e1666. 254 Kinnaert E, Duez P, Morandini R, Dubois J, Van Houtte P, 247 Minchinton AI, Chaplin DJ. Spatial characterization of gluta- Ghanem G. Cysteine but not glutathione modulates the thione depletion in the KHT sarcoma using flow cytometry. Int radiosensitivity of human melanoma cells by affecting both J Radiat Biol 1991;59:1425e1433. survival and DNA damage. Pigment Cell Res 2004;17:275e280. 248 Vukovic V, Nicklee T, Hedley DW. Microregional heterogeneity 255 Biaglow JE, Ayene IS, Koch CJ, et al. Radiation response of of non-protein thiols in cervical carcinomas assessed by cells during altered protein thiol redox. Radiat Res 2003;159: combined use of HPLC and fluorescence image analysis. Clin 484e494. Cancer Res 2000;6:1826e1832. 256 O’Neill P. Pulse radiolytic study of the interaction of thiols and 249 Vukovic V, Nicklee T, Hedley DW. Differential effects of ascorbate with OH adducts of dGMP and dG: implications for buthionine sulphoximine in hypoxic and non-hypoxic regions of DNA repair processes. Radiat Res 1983;96:198e210. human cervical carcinoma xenografts. Radiother Oncol 2001; 257 Hodgkiss RJ, Stratford MRL. Competitive dose-modification 60:69e73. between ascorbate and misonidazole in human and hamster 250 Bailey HH, Ripple G, Tutsch KD, et al. Phase I study cells: effects of glutathione depletion. Int J Radiat Biol 1988; of continuous-infusion L-S,R-buthionine sulfoximine with 54:601e610. intravenous melphalan. J Natl Cancer Inst 1997;89:1789e1796. 258 Coleman CN, Mitchell JB. Clinical radiosensitization: why it 251 Bailey HH. L-S,R-buthionine sulfoximine: historical develop- does and does not work. J Clin Oncol 1999;17:1e3. ment and clinical issues. Chem Biol Interact 1998;111e112: 259 Hornig D. Distribution of ascorbic acid, metabolites and ana- 239e254. logues in man and animals. Ann NY Acad Sci 1975;258:103e117. 252 Hamilton D, Wu J, Batist G. Structure-based identification of 260 Wardman P. Evaluation of the ‘radical sink’ hypothesis from novel classes of human g-glutamylcysteine synthetase a chemical-kinetic viewpoint. J Radioanal Nucl Chem 1998; inhibitors. Mol Pharmacol 2007;71:1140e1147. 232:23e27.